Dissertations / Theses on the topic 'Predikcija'

U radu je prikazan model podataka koji omogućavapredviđanje uspješnosti studiranja na visokoškolskimustanovama, kao i analizu više tehnika predikcije.Pored toga, prikazuje i prototipsku implementacijuinformacionog sistema za upravljanje obrazovnimprocesom koji omogućava korišćenje predikcije urealnim informacionim sistemima.
The paper presents a data model that facilitatesprediction of students success in studying, as well asa review of prediction techniques. It also presents aprototype implementation of a learning managementinformation system that enables the use of predictionof success in studying and represents a real-worlduse case.

Ekstremni vidovi ispoljavanja agresije u vidu krivičnih dela nasilja su univerzalnifenomeni, prepoznati u svim društvima i kulturama. Iako se radi o relativno nefrekventnimdogađajima, trend nasilničkog ponašanja raste i predstavlja problem od šireg društvenog značaja.Ishod višedecenijskog izučavanja agresivnosti ogleda se u detektovanju velikog broja prediktoraagresivnog ponašanja, pri čemu su se faktori agresivnosti najčešće izučavali izolovano. Kaoreferentni okvir za ovo istraživanje poslužio nam je Opšti model agresivnosti koji podrazumevameđusobnu interakciju različitih faktora u generisanju agresivnog ponašanja. Osnovni cilj ovogistraživanja se odnosi na rasvetljavanje glavnih i interaktivnih efekta prediktora, a što doprinosiboljem razumevanju uslova pod kojima se povećava ili smanjuje verovatnoća realizacijeagresivnog ponašanja, u kontekstu krivičnih dela.Istraživanjem je obuhvaćeno 200 ispitanika, muškog pola, podeljenih u dve grupe.Kriterijsku grupu činilo je 100 ispitanika koji su bili na izdržavanju kazne u Kazneno popravnomzatvoru u Sremskoj Mitrovici, zbog krivičnog dela nasilja. Kontrolnu grupu činilo je 100ispitnanika koji u svojoj istoriji nisu imali izvršeno ni jedno krivično delo. Ispitanici suujednačeni u odnosu na psihijatrijsku dijagnozu.Organizovan je korelacioni nacrt, a rezultati su obrađeni kroz transferzalnu perspektivu.Ispitivanje interaktivnog uticaja prediktorskih varijabli sprovedeno je putem ispitivanjamoderacije.Rezultati istraživanja pokazuju da je na osnovu personološko-dispozicionih, kognitivnoemocionalnihi socio-emografskih prediktora moguće razlikovati grupu počinilaca krivičnog delanasilja u odnosu na ispitanike koji nikada nisu počinili krivično delo. Konkretnije, grupupočinilaca krivičnih dela nasilja karakteriše viši stepen izraženosti sve tri Eysenck-ovedimenzije, viši stepen sklonosti ka agresivnom ponašanju, viši stepen izraženosti psihopatskedevijacije, dok se po pitanju stepena samopoštovanja ne razlikuju u odnosu na kontrolnu grupu.Počinioci krivičnog dela nasilja imaju specifičan kognitivni stil koji je obeležen većimprisustvom agresivnih fantazija, neefikasnom kontrolom agresivnih i neprijatnih misli, većimprisustvom iracionalnih uverenja, uz češće korišćenje supresije, kao neadekvatne strategijeviemocionalne regulacije. Takođe, počinioci krivičnog dela potiču iz porodica sa većim stepenomalkoholizma (isključivo oca), u većoj meri su bili izloženi nasilju u formativnom periodu, imajuniži stepen obrazovanja i ređe imaju stalno zaposlenje. Međutim, kada se ovi brojni faktoriagresivnosti posmatraju kroz prizmu multivarijatnog modela predikcije, samo mali broj ostvarujeglavni prediktivni doprinos. Izdvojili su se sledeći prediktori: sklonost ka antisocijalnomponašanju, samopoštovanje, netolerancija životnih događaja, supresija, reprocenjivanje iobrazovni status. Ispitujući interaktivan efekat prediktorskih varijabili i psihopatije, kaomoderator varijable, rezultati pokazuju da različit stepen izraženosti subdimenzija psihopatijepredstavlja uslov pod kojim personološko-dispozicione varijable ostvaruju značajan doprinos uprdikciji agresivnog ponašanja.Uzimajući u obzir da su se kognitivno-emocionalni faktori koji su podložni promeniizdvojili kao značajni prediktori, praktičan cilj istraživanja ogleda se u primeni nalazaistraživanja na proces rehabilitacije počinilaca agresivnih krivičnih dela
Extreme forms of aggression manifestations, in terms of violent crimes, are universalphenomena recognized in all societies and cultures. Although these are relatively smallfrequency events, the trend of violent behaviour is growing and represents a problem of widersocial significance. The result of multiple decades researches of human aggression is thedetection of a large number of aggressive behaviour predictors, where the aggression factorswere most commonly studied as isolated ones. As a reference framework for this research, theGeneral Aggression Model was used, as it comprehends different factors mutual interaction ingeneration of aggressive behaviour. The basic aim of this research is to put some more light tothe main and interactive predictor effects, which contributes to better understanding of theconditions under which the probability of realization of the aggressive behaviour is rising orlowering, in terms of criminal acts.The research was performed on 200 male subjects divided into two groups. The criteriagroup was formed out of 100 subjects who are imprisoned in Sremska Mitrovica Penitentiary forviolent crimes. The control group was formed out of 100 subjects who do not have a criminalhistory at all. The subjects are uniform with relation to psychiatric diagnosis.Correlation design was organized and the results were processed through transversalperspective. Examination of the interactive influence of the predictor variables was performedthrough moderation.Research results show that it is possible to distinguish the group of violence offendersfrom the group of subjects with no criminal history at all, on the basis of personologicaldispositional,cognitive-emotional and socio-demographical predictors. Specifically, the group ofviolent offenders is characterized by the higher level of expression of all three Eysenckdimensions, higher level of inclination to aggressive behaviour, higher level of expression ofpsychopathic deviation, while the level of self-esteem is no different to the control group.Violent offenders have a specific cognitive style which is marked by higher presence ofaggressive fantasies, non-efficient control of aggressive and unwanted thoughts, higher presenceof irrational beliefs, with more frequent use of suppression as inadequate strategy for emotionalviiiregulation. Also, violent offenders come from families with higher level of alcoholism (fatheronly), they have been exposed to violence to a bigger extent in their formation period, they havea lower level of education and less frequently have a permanent employment. However, whenthese numerous factors of aggression are observed through the prism of multivariate model ofprediction, only a small number of factors realize the main predictive effect. The followingpredictors are noted as significant: inclination to antisocial behaviour, self-esteem, lowfrustration tolerance beliefs, suppression, reappraisal and educational status. Examination ofinteractive effect of predictor variables and psychopathy, as moderator variable, gives resultswhich show that different degree of expression of psychopathy sub-dimensions represents thecondition under which the personological-dispositional variables give significant contribution toaggressive behaviour prediction.Taking into account that the cognitive-emotional factors which are subject of change areshown to be significant predictors, the practical aim of this research is to apply the researchresults in violent offenders’ rehabilitation process.

The topic of the research in this dissertation is the application of machinelearning in solving problems characteristic to biosystems, with specialemphasis on agriculture. Firstly, an innovative regression algorithm based onbig data was presented, that was used for yield prediction. The predictionswere then used as an input for the improved portfolio optimisation algorithm,so that appropriate soybean varieties could be selected for fields withdistinctive parameters. Lastly, a multi-objective optimisation problem was setup and solved using a novel method for categorical evolutionary algorithmbased on NSGA-III.
Предмет истраживања докторске дисертације је примена машинског учења у решавању проблема карактеристичних за биосистемe са нагласком на пољопривреду. Најпре је представљен иновативни алгоритам за регресију који је примењен на великој количини података како би се са предиковали приноси. На основу предикција одабране су одговарајуће сорте соје за њиве са одређеним карактеристикама унапређеним алгоритмом оптимизације портфолија. Напослетку је постављен оптимизациони проблем одређивања сетвене структуре са вишеструким функцијама циља који је решен иновативном методом, категоричким еволутивним алгоритмом заснованом на NSGA-III алгоритму.
Predmet istraživanja doktorske disertacije je primena mašinskog učenja u rešavanju problema karakterističnih za biosisteme sa naglaskom na poljoprivredu. Najpre je predstavljen inovativni algoritam za regresiju koji je primenjen na velikoj količini podataka kako bi se sa predikovali prinosi. Na osnovu predikcija odabrane su odgovarajuće sorte soje za njive sa određenim karakteristikama unapređenim algoritmom optimizacije portfolija. Naposletku je postavljen optimizacioni problem određivanja setvene strukture sa višestrukim funkcijama cilja koji je rešen inovativnom metodom, kategoričkim evolutivnim algoritmom zasnovanom na NSGA-III algoritmu.

Примарни задатак ове докторске дисертације је развој предиктивногмодела фреквенције саобраћајних незгода на руралним путевима.Секундарни циљ је процјена ефеката саобраћајних и геометријскихваријабли на фреквенцију саобраћајних незгода. Резултатиистраживања су показали да је најбоље перформансе, у погледуслагања са стварним подацима и предвиђања саобраћајних незгода, уодносу на друге конкурентске моделе показао НБ модел. Постављеноограничење брзине има највећи релативни утицај на фреквенцијуукупног броја саобраћајних незгода и саобраћајних незгода санастрадалим лицима. На фреквенцију саобраћајних незгода саматеријалном штетом највећи релативни утицај има дужина сегмента.
Primarni zadatak ove doktorske disertacije je razvoj prediktivnogmodela frekvencije saobraćajnih nezgoda na ruralnim putevima.Sekundarni cilj je procjena efekata saobraćajnih i geometrijskihvarijabli na frekvenciju saobraćajnih nezgoda. Rezultatiistraživanja su pokazali da je najbolje performanse, u pogleduslaganja sa stvarnim podacima i predviđanja saobraćajnih nezgoda, uodnosu na druge konkurentske modele pokazao NB model. Postavljenoograničenje brzine ima najveći relativni uticaj na frekvencijuukupnog broja saobraćajnih nezgoda i saobraćajnih nezgoda sanastradalim licima. Na frekvenciju saobraćajnih nezgoda samaterijalnom štetom najveći relativni uticaj ima dužina segmenta.
The primary objective of this doctoral dissertation is the development of theaccident frequency predictive model on rural roads. The secondary objectiveis to evaluate the effects of traffic and geometric variables on the accidentfrequency. The results of the research showed that the best performance, interms of goodness-of-fit measures and predicting accidents, was shown bythe NB model in comparison with other competing models. Posted speedlimit has the highest relative influence on the accident frequency of the totalnumber of accidents and the fatal and injury accidents. The independentvariable segment lenht has the greatest relative influence on the accidentswith property damage frequency.

The thesis is devoted to the position of rating agencies in the financial markets. The main aim is to predict the position of these companies, which is currently mainly influenced by the European politicians. Politicians criticize them and try to limit their power. The essence of this thesis is to summarize the reasons for their criticism, focusing on reduction of sovereign ratings. Then it's discussed a downgrading the USA and France. For the first time in a history they lost their AAA rating. The last chapter is devoted to the possible position of rating agencies. The first option is the establishment of the European rating agency, the other one is tightening regulation and the last is a situation if the rating agencies were abolished.

Stalno praćenje rada softvera je neophodno da bi se utvrdilo da li softver poštuje zadate nivoe kvaliteta. Na osnovu sakupljenih podataka, moguće je da se predvidi i dalje ponašanje aplikacije i da se izvrši izbor daljih akcija da bi se održao zahtevani nivo. Tema ove disertacije je razvoj sistema za kontinualno praćenje performansi softvera, kao i razvoj modela za predviđanje performansi softvera. Za implementaciju sistema potrebljena je JEE tehnologija, ali je sistem razvijen tako da može da se primeni i za praćenje softvera razvijenog za druge platforme. Sistem je modelovan tako minimalno utiče na performanse sistema softvera koji se prati. Linearna regresija je upotrebljena za modelovanje zavisnosti performansi od okruženja u kom se softver izvršava. Sistem je upotrebljen za praćenje izabrane JEE aplikacije.
Continuous monitoring of software is necessary to determine whether the software performs within required service perfomance levels. Based on collected data, it is possible to predict the future performance of applications and to plan further actions in order to maintain the required service levels. The theme of this dissertation is the development of systems for continuous performance monitoring software, as well as the development of models for predicting the performance of software. To implement the system was used JEE technologies, but the system was developed so that it can be used for tracking software developed for other platforms. The system is modeled as a minimum impact on system performance software that is monitored. Linear regression was used for modeling the dependence of the performance environment in which the software is running. The system was used to monitor selected JEE applications.

UVOD Prenatalna dijagnostika predstavlja skup metoda i postupaka čiji je cilj da potvrde ili isključe postojanje kongenitalnih anomalija ploda. Prenatalni skrining može biti ne invazivni i invazivni. Ne invazivni skrining treba da ima visku senzitivnost i da omogući adekvatnu selekciju trudnica kojima će se predložiti genetsko ispitivanje ploda iz uzoraka dobijenih invazivnim metodama prenatalne dijagnostike. Prenatalni skrining prvog trimestra trudnoće obuhvata ultrazvučni pregled debljine nuhalne translucencije i laboratorijsku analizu dva biohemijska markera od 11 do 14 nedelje trudnoće, Prenatalni skrining drugog trimestra trudnoće koji se zasniva na biohemijskom skriningu i tripl testu iako je jedini koji se primenjuje zbog niske senzitivnosti od 20% do 40%, ne može se smatrati validnim. Integrativni biohemijski test prvog i drugog trimestra imaj veću senzitivnost (od 40 do 60%) ali ni on nije dao očekivane rezultate u adekvatnoj selekciji trudnica za genetsku analizu ploda zbog visoke stope lažno pozitivnih rezultata. Drugi trimestar trudnoće omogućava sonografskim pregledima i biohemijskim analizama dopunski način a u nekim slučajevima i jedini u proceni postojanja rizika Daunovog sindroma ili nekih drugih hromozomskih aberacija ploda Zato je primena integrativnih prenatalnih ne invazivnih metoda prvog i drugog trimestra trudnoće veoma značaja u poboljšanju dijagnostičkih vrednosti prenatalnih skrining testova i ima za cilj da smanji procenat invazivnih procedura zbog mogućih komplikacija i ne potrebnih finansijskih troškova. Daunov sindrom(trizomija 21 para hromozoma) je najčešća hromozomska numerička aberacija praćena mentalnom retardacijom dece (I.Q<70. ) Deca sa Daunovim sindromom su karakterističnog fenotipskog izgleda i sa čestim kongenitalnim anomailjama koje im onemogućavaju normalan život a često su i uzrok njihove prerane smrtnosti. Kongenitalne anomalije su zastupljene kod 2% do 5% živo rođene dece, predstavljaju 25 % perinatalne smrtnosti, četvrtina su uslovljnene hromozomskim aberacijama ili naslednom osnovom, od čega 0, 2%-0, 4% su sa Daunovim sindromom. CILJEVI Ciljevi četrorogodišnjeg istraživanja su bili da se poboljša dijagnostička vrednost postojećih prenatalnih testova, da se potvrdi značaj ultrazvučnog skrininga drugog trimestra trudnoće analizom debljine vratne brazde i dužine butne kosti ploda te da se poboljša njegova senzitivnost korporativnom sonografskom analizom cefaličnog indexa, intraorbitalnog rastojanja i dužine fronto-talamične distance. MATERIJAL I METODE Ukupan broj trudnica obuhvaćen četvorogodišnjim ispitivanjem koje su ultrazvučno pregledane i kojima je savetovano genetsko ispitivanje ploda blio je 4552. Tokom Retrospektivnog dvogodišnjieg ( 2010.2011)bila je 2169 dok je prospektivnom dvogodišnjom analizom (2012, 2013)je bilo obuhvaćeno 2383 trudnica. Ispitivana grupa su bile trudnice kod kojih je genetskom analizom otkriven patološki kariotip ploda, kontrolna grupa je obuhvatila sve ostale trudnice kod je kariotip ploda bio normalan od kojih su 124 trudnice odabrane metodom slučajnog izbora. Retrospektivnom studijom ultrazvučna je pregledana dužina vratne brazde(>6mm i dužina butne kosti<0, 6 od 14 do 22 nedelje trudnoće. Analizirana je cirkulacija fetalne krvi kroz duktus venosus ploda( a talas) i postojanje nosne kosti ploda(+, -). Prospektivnom analizom je ultrazvučnim pregledom ploda dodatno analiziran cefalični index(>85%), i intraorbitalna distanca i duzina fronto-talamične distance(<80%) ploda. Korišćene su metode deskriptivne statističke analize, aritmetička sredina, standardna devijacija, najmanja i najveća vrednost kod parametrijskih obeležja dok su za ne parametrijska postojanje nosne kosti i alfa talasa duktusu venozusu korišćene druge statističke metode, a komparativnim statističkim metodama kod normalnih, patoloških i kariotipova sa Daunovim sindromom ploda. Statistička značajnost je dokazana t testom a definisana nivoom p<0, 05 i p<0, 001 odnosom kod normalnih, patoloških kariotipova i Daunovog sindroma. Multifaktorskom regresivnom logističkom analizom je urađena procena senzitvnosti prenatalnog ultrazvučnog skrininga sa ispitivanim obeležjima drugog trimestra trudnoće REZULTATI I DISKUSIJA Od ukupnog broja ultrazvučno pregledaninh trudnica 4552 kojima je savetovano genetska analiza ploda citogenetskom analizom je otkriveno 66 patoloških kariotipova 1, 49%, sa Dunovim sindromom 31 0, 68%. Deskriptivnom statističkom obradom ultrazvučno ispitivanih obeležja od 14 do 22 nedelje trudnoće, uočeno je odstupanje i potvrđen značaj starije životne dobi trudnica, debljine vratne brazde i dužine frontotalamične distance u odnosu na normalne nalaze katiotipova ploda u predikciji Daunovog sindroma.Vrednosti dužine butne kosti, cefaličnog indeksa i intraorbitalnog rastojanja nisu imala veća odstupanja u poređenju patoškokih i normalnih nalaza kariotipova.Studentovim t testom je i dokazano p<0, 001 za debljinu vratne brazde i dužinu fronto-talamične distance, dok je za stariju životnu dob trudnice potvrđeno a;0, 001. Senzitivnost prenatalnog skrininga drugog trimestra analizom debljine vratne brazde i dužine butne kosti je veća u odnosu na standardno primenjivan biohemijski skrining drugog trimestra tripl testa (senzitivnost 40%-60) sa velikom stopom lažno pozitivnih rezultata.Dokazan je značaj poboljšanja senzitivnosti prenatalnih skrining testova dopunskom analizom tri ultrazvučna parametra, dužine fronto-talamične distance, cefaličnog indeksa i intraorbitalnog rastojanja u predikciji Daunovog sindroma, ali i kod ostalih hromozomskih aberacija ploda u periodu od 14 do 22 nedelje trudnoće primenomi multifaktorske logističke regresivne analize senzitivnost preko 93% sa 7% lažno pozitivnih rezultata. Postojanje korelacije između debljine vratne brazde i dužine fronto-talamične distance ploda poboljšavai senzitivnost prenatalnih ultrazvučnog skrininga. Integrativnim pristupom ultrazvučnog i biohemijskog skrininga drugog trimesra trudnoće, tripl testa očekuje se poboljšati dijagnostičkih vrednosti prenatalnog skrininga senzitivnost ne invazivnog skrininga u predikciji Daunovog sindroma i ostalih hromozomskih aberacija ploda. ZAKLJUČCI 1Potvrđen je značaj starije životne dobi trudnice u povećanju rizika Daunovog sindoma, i ostalih hromozomskih aberacija ploda ( p<0, 001) Potvrđen je značaj zadebljanja vratne brazde ploda >6mm(p<0, 001) i skraćenja butne kosti kod Daunovog sindroma ploda od 14 do 22 nedelje trudnoće u prenatalnom otkrivanju Dunovog sindroma i ostalih hromozomskih aberacija ploda i selekciji trudnica kojima će se predložiti genetsko ispitivanje ploda.Potvrđena je hipoteza da skraćenje fronto-talamične distance poboljšava senzitivnost ultrasonografskog skrininga, jer češće postoji kod Daunovog sindroma ploda ali i ostalih numeričkih hromozomskih aberacija tipa, nego kod normalnih nalaza kariotipa ploda( p<0, 001).Komparativnom analzom ultrazvučnim pregledom fronto-talamična distance debljine vratne brazde i dužine butne kosti ploda od 14 do 22 nedelje trudnoće može se značajno poboljšati vrednost dijagnostičkih prenatalnih testova u predikciji Daunovog sindroma. Postojanje korelacija između fronto-talamične distance i debljine vratne brazde dopunjuje ultrazvučni skrining i povećava njegovu senzitivnost na preko 90%, što je multifaktorskom regresivnom logaritamskom analizom i potvrđeno. Značaj multidisciplinarnog pristupa pogotovo je izražen u predikciji Daunovog sindroma, obzirom na različite specijalnosti koje u njemu učestvuju. Cost – benefit analiza. Visoka senzitivnost ne invazivnog prenatalnog skrininga u predikciji Daunovog sindroma, smanjuje troškove za pojedince i državu jer je njihova cena i do dest puta manja od cene citogenetskih analiza, a i trudnice se ne izlažu riziku mogućih komplikacija prilikom izvođenja invazivnih metoda
INTRODUCTIONS Prenatal diagnostic procedure represent a set of methods and techniques with the aim to afirmate or eliminate the presence of Down’s syndrome and other congenital anomalies Can be non-invasive and invasive methods. Non-invasive methods (laboratory or ultrasonographic) have the aim to make possible the most valid assessment of the risk of presence of an affected fetus in the pregnancy, selected pregnancy for invasive diagnostics procedures and citogenetics analisseskariotipingfoeti. Down’s syndrome, aneuploidy with trisomy 21 chromosomal, is the most common chromosomal numerical aberration associated with mental retardation of children (IQ< 70). Children with Down’s syndrome have characteristic phenotypic appearance with high frequent congenital anomalies that preclude a normal life and are frequently the cause of their earlier death. AIM The aim of the four year long investigation was to confirm the importance of ultrasound screening by the analyses of the basic ultrasound parameters for the second trimester, the thickness of the nuchal fold and the length of the femur of the fetus in the prediction of Down’s syndrome and other chromosomal aberrations of the fetus, as well as to improve other existing ultrasonic screenings of the first and second trimester of pregnancy by ultrasonic examination and analyses of the cephalic index and intraorbital space and the length of the fronto-thalamic distance. MATERIAL AND METODS Retrospective investigation (2010. 2011) and prospective investigation (2012.2013) includes 4655 pregnant women. For all pregnant women the genetic investigation of the fetus was performed. A total of 68 were found with chromosomal aberrations, 38 with Down’s syndrome. The method of haphazard choice in retrospective study and in prospective study ultrasound markers are examined. In retrospective analyses of the nuchal fold (<6mm and the length of femur <0.6, that represent basic ultrasound screening of the second trimester and are analyzed as parametric signs of the second trimester, and are analyzed as parametric markers, and analyses of the circulation of fetal blood through ductus venosus of the fetus. In the retrospective study the length of the nuchal fold (>6mm in length, that represent a basic ultrasound screening of the second trimester, and are analyzed as parametric markers in the prediction of Down’s syndrome and other chromosomal aberrations. RESULTS AND DISCUSION Cytogenetic analyses revealed 66 (1, 49%) pathologic karyotypes and Down syndrome were present in 31 (0, 68%) cases. All pathologic karyotypes were obtained after ultrasound examinations of 4552 pregnant women. Ultrasound markers for period 14th-22nd GW were analyzed with descriptive statistical methods and importance of pregnancy in older women, thickness of nuchal fold and lengths frontal thalamic distance were proofed in case of Down syndrome. Femoral bone lengths, cephalic index and intraorbital distances were similar for both groups, normal and pathologic karyotypes. Student’s t test revealed statistical significance with p<0, 001 values for nuchal fold thickness, frontal thalamic distance and older ages.Three additional ultrasound markers (frontal thalamic distance, cephalic index, intraorbital distance) improve prediction of Down syndrome and other chromosomal aberrations between 14th and 22nd GW as well. Multifactorial logistic regressive analyses revealed 93% sensitivity with 7% false positive results. Corelation between nuchal fold thickness and frontal thalamic distance improve prenatal ultrasound screening sensitivity. Using both ultrasound and biochemical screening (triple test) is way to improve sensitivity of non invasive screening in prediction of Down syndrome and other chromosomal aberrations. CONCLUSIONS Importance of pregnant women ages and higher risk for Down syndrome and other chromosomal aberrations was proofed (p<0, 001).Importance of nuchal fold thickness above 6mm (p<0, 001) and shorter femoral bone marker in period from 14th to 22nd GW in prediction of Down syndrome and other chromosomal aberrations are proofed (p<0, 001). Hypothesses that frontal thalamic distance improve ultrasound screening sensitivity was proofed was proofed (p<0, 001) since it is significantly shorter in Down syndrome and other chromosomal aberrations in comparison with fetuses with normal karyotypes. Comparative analyses of frontothalamic distance, nuchal fold thickness and femoral bone length in period from 14th to 22nd GW can signifi cantly improve prenatal diagnostic testing in Down syndrome prediction. Correlation between frontothalamic distance and nuchal fold thickness improve ultrasound screening sensitivity on 93% that is proofed with multifactorial logistic regressive analyses. Significance of multidisciplinary approach is high in Down syndrome prediction. Cost-benefit: High sensitivity of non invasive prenatal screening in Down syndrome prediction reduces costs for families and government since it costs ten time less than cytogenetic analyses and risk with invasive procedures is avoided.

Master thesis deals with possibilities of predicting traffic situation on the macroscopic level using data, that were recorded using traffic sensors. This sensors could be loop detectors, radar detectors or cameras. The main problem discussed in this thesis is the travel time of cars. A method for travel time prediction was designed and implemented as a part of this thesis. Data from real traffic were used to test the designed method. The first objective of this thesis is to become familiar with the prediction methods that will be used. The main objective is to use the acquired knowledge to design and to implement an aplication that will predict required traffic variables.

This master's thesis addresses the solubility of recombinant proteins and its prediction. It describes the subject of protein synthesis, as well as the process of recombinant protein creation. Recombinant protein synthesis is of great importance for example to pharmacologic industry. This synthesis is not a simple task and it does not always produce viable proteins. Protein solubility is an important factor, determining the viability of the resulting proteins. It is of course favourable for companies, that take part in recombinant protein synthesis, to focus their effort and their resources on proteins, that will be viable in the end. In this regard, bioinformatics is of great help, as it is capable, with the help of machine learning, of predicting the solubility of proteins, for example based on their sequences. This thesis introduces the reader to the basic principles of machine learning and presents several machine learning methods, used in the field of protein solubility prediction. It deals with the definition of a dataset, which is later used to test selected predictors, as well as to train the ensemble predictor, which is the main focus of this thesis. It also focuses on several specific protein solubility predictors and explains the basic principles upon which they are built, as well as the results of their testing. In the end, it presents the ensemble predictor of protein solubility.

The goal of this thesis is test and comparation of the offline tools for prediction of protein structure and creation of metaprediktor, which allows the user to select the appropriate tool, according to given parameters. Testing tool is based on a dataset of proteins, which is based on the SCOP database and it is trying to be as balanced as possible to include proteins from different families and thus could best evaluate individual tools. The results of this thesis are requirements of metaprediktor and also which data and settings can be allowed and processed and how it will be implemented.

This thesis aims to optimize the machine learning algorithms for predicting KPI metrics for an organization. The organization is predicting whether projects meet planned deadlines of the last phase of development process using machine learning. The work focuses on the analysis of prediction models and sets the goal of selecting new candidate models for the prediction system. We have implemented a system that automatically selects the best feature variables for learning. Trained models were evaluated by several performance metrics and the best candidates were chosen for the prediction. Candidate models achieved higher accuracy, which means, that the prediction system provides more reliable responses. We suggested other improvements that could increase the accuracy of the forecast.

Porozumění chování silniční dopravy je klíčem pro její efektivní řízení a organizaci. Tato úloha se stává čím dál více důležitou s rostoucími požadavky na dopravu a počtem registrovaných vozidel. Informace o dopravní situaci je důležitá pro řidiče a osoby zodpovědné za její řízení. Naštěstí v posledních několika dekádách došlo k značnému rozvoji technologií pro monitorování dopravní situace. Stacionární senzory, jako jsou indukční smyčky, radary, kamery a infračervené senzory, mohou být nainstalovány na důležitých místech. Zde jsou schopny měřit různé mikroskopické a makroskopické dopravní veličiny. Bohužel mnohá měření obsahují nekorektní data, která není možné použít při dalším zpracování, například pro predikci dopravy a její inteligentní řízení. Tato nekorektní data mohou být způsobena poruchou zařízení nebo problémy při přenosu dat. Z tohoto důvodu je důležité navrhnout obecný framework, který je schopný doplnit chybějící data. Navíc by tento framework měl být také schopen poskytovat krátkodobou predikci budoucího stavu dopravy. Tato práce se především zabývá vybranými problémy v oblasti doplnění chybějících dopravních dat, predikcí dopravy v krátkém časovém horizontu a predikcí dojezdových dob. Navrhovaná řešení jsou založena na kombinaci současných metod strojového učení, například Support vector regression (SVR) a multikriteriálních evolučních algoritmů. SVR má mnoho meta-parametrů, které je nutné dobře nastavit tak, aby byla dosažena co nejkvalitnější predikce. Kvalita predikce SVR dále silně závisí na výběru vhodné množiny vstupních proměnných. V této práci používáme multiktriteriální optimalizaci pro optimalizaci SVR meta-parametrů a množiny vstupních proměnných. Multikriteriální optimalizace nám umožňuje získat mnoho Pareto nedominovaných řešení. Mezi těmito řešeními je možné dynamicky přepínat dle toho, jaká data jsou aktuálně k dispozici tak, aby bylo dosaženo maximální kvality predikce. Metody navržené v této práci jsou především vhodné pro prostředí s velkým množstvím chybějících hodnot v dopravních datech. Tyto metody jsme ověřili na reálných datech a porovnali jejich výsledky s metodami, které jsou v současné době používány. Navržené metody poskytují lepší výsledky než stávající metody, a to především ve scénářích, kde se vyskytuje mnoho chybějících hodnot v dopravních datech.

This work thoroughly analyzes a previously created computational model for predicting characteristic properties of the combustion process in an experimental combustion chamber. Any found shortcomings of the original model are removed and the model is further improved prior to its application on 11 real cases of combustion tests performed at various conditions and with various fuels. Data provided by the model are confronted with the data obtained during the combustion tests and the model accuracy is evaluated, based on local heat flux along the length of the combustion chamber. Finally, the overall usefulness of the model is determined by the means of evaluating the acquired accuracy values, and further possibilities of model improvement and use are presented.

This master’s thesis deals with protein secondary structure prediction. There is a theoretical introduction followed by study of available tools, proposal and implementation of web application, which combines functionality of several web tools used to predict secondary structure. User is asked to choose prediction methods and insert input sequence as plain text or upload a file. Results collected from selected tools serve to convert data into common format, show the result and create new type of prediction. Finally, the testing is applied and influences of tools are adjusted in order to increase percentage of prediction. The output of application is a result of prediction also available as plain text or as a file.

This work deals with proteins, especially with their structure and kinds of tertiary, or 3D, structure prediction. Tertiary structure prediction is very important for function prediction of this vitally important substance. Bioinformatics do this prediction much more effective and faster, because classical methods of structure prediction directly from molecule are very expensive and slow. On the other hand they are much more exact. Objective of this thesis is to describe tertiary structure prediction methods, describe used tools and possibility of automatic communication with them.  Next objective is describe implementation of server, that will serve to protein engineers for more effective finding of information about tertiary structure from more servers without requesting each of them separately. Results of testing will be described in this work too.

The aim of this study is to evaluate feasibility of using Bayes classifiers for predicting corporate bankruptcies. The results obtain show that Bayes classifiers do reach comparable results to then more commonly used methods such the logistic regression and the decision trees. The comparison has been carried out based on Czech and Polish data sets. The overall accuracy rate of these so called naive Bayes classifiers, using entropic discretization along with the hybrid pre-selection of the explanatory attributes, reaches 77.19 % for the Czech dataset and 79.76 % for the Polish set respectively. The AUC values for these data sets are 0.81 and 0.87. The results obtained for the Polish data set have been compared to the already published articles by Tsai (2009) and Wang et al. (2014) who applied different classification algorithms. The method proposed in my study, when compared to the above earlier works, comes out as quite successful. The thesis also includes comparing various approaches as regards the discretisation of numerical attributes and selecting the relevant explanatory attributes. These are the key issues for increasing performance of the naive Bayes classifiers

Dizertační práce se zabývá predikcí vysokodimenzionálních dat genových expresí. Množství dostupných genomických dat významně vzrostlo v průběhu posledního desetiletí. Kombinování dat genových expresí s dalšími daty nachází uplatnění v mnoha oblastech. Například v klinickém řízení rakoviny (clinical cancer management) může přispět k přesnějšímu určení prognózy nemocí. Hlavní část této dizertační práce je zaměřena na kombinování dat genových expresí a klinických dat. Používáme logistické regresní modely vytvořené prostřednictvím různých regularizačních technik. Generalizované lineární modely umožňují kombinování modelů s různou strukturou dat. V dizertační práci je ukázáno, že kombinování modelu dat genových expresí a klinických dat může vést ke zpřesnění výsledku predikce oproti vytvoření modelu pouze z dat genových expresí nebo klinických dat. Navrhované postupy přitom nejsou výpočetně náročné.  Testování je provedeno nejprve se simulovanými datovými sadami v různých nastaveních a následně s~reálnými srovnávacími daty. Také se zde zabýváme určením přídavné hodnoty microarray dat. Dizertační práce obsahuje porovnání příznaků vybraných pomocí klasifikátoru genových expresí na pěti různých sadách dat týkajících se rakoviny prsu. Navrhujeme také postup výběru příznaků, který kombinuje data genových expresí a znalosti z genových ontologií.

Práce se zabývá popisem technické analýzy jako jedné z částí celkové analýzy instrumentu. V práci jsou detailně popsány nejpoužívanější technické indikátory a grafické formace. Autor se v práci zamýšlí nad vznikem technické analýzy a odhaduje její možnosti použití do budoucna. V na akciích Burzy cenných papírů Praha a.s. potom zkoumá úspěšnost predikcí technické analýzy.

Osnovni doprinos ove doktorske disertacije je razvoj algortima i sistema za objektivnu procenu vizuelnog kvaliteta slike uzimajući u obzir najvažnija moguća oštećenja kao što su zamućenje ivica (oštrina) i poremećaj prirodnog izgleda teksture objekata na slici sa jedne strane i uticaj sadržaja slike (procenta ivica u slici) na procenu kvaliteta sa druge strane. Dakle, hipoteza izneta u ovom radu je da je potreban multiparametarski pristup da bi se dobila objektivna procena kvaliteta slike koja je što približnija subjektivnoj proceni.

This thesis discusses travel time prediction of vehicles on roads based on the methods of machine learning. It describes theory of travel times and summarizes scientific papers dealing with this topic. Within the thesis, analysis of real travel time data was done and the features to be used in prediction models were engineered. Finally, the complex prediction system was designed and implemented and has been tested in production environment.

This work deals with user satisfaction with mobile data service. The main part is devoted to the development of a mobile application that provides testing network parameters to evaluate and use online databases shared with other users. Discussed the possibility of distributing applications among users over its testing and then to the App Store.

The doctoral thesis deals with the finding of a suitable approach for predicting the development of the residential real estate market, which would also be applicable in the practice of Experts and further develop the appraisal field. It has been found that a delayed multiple linear regression model appears to be appropriate, as confirmed by the verification of this model. The resulting model is also suitable for use in routine Expert practice, thanks to the simplicity of calculation without ownership of any computing program. Expert thanks to the created model is able to predict the development of the real estate market. The result is bound to the accuracy of the input data. All assumptions of regression models have been tested, optimal explanatory variables were selected based on backword regression. The doctoral thesis explains all input data, methods, tests, procedures and detailed modeling.

This work deals with problem of analyzing cooling curves of ductile iron and discovering possibility to predict shrinkage tendency of melt from thermal analysis results. This work was executed in foundry production conditions where obtained data from thermal analysis data linked to experimental castings where amount of shrinkage was estimated. Data were analyzed and looked for correlations between acquired thermal data and shrinkage defects.

Uvod: Pravovremena i pouzdana stratifikacija rizika pacijenata sa akutnim ST eleviranim infarktom miokarda (STEMI) je važna zbog adekvatnog zbrinjavanja ovih bolesnika. Primarna perkutana koronarna intervencija (pPCI) je dovela do značajnog poboljšanja ishoda pacijenata sa STEMI, a time moguće i do promene prediktivne važnosti različitih faktora rizika. Jung variabla je jednostavan klinički indeks rizika koji se koristi tokom inicijalne prezentacije pacijenata. Pokazana je prediktivna vrednost Jung varijable za hospitalni mortalitet pacijenata sa STEMI lečenih fibrinolitičkom terapijom, dok uloga Jung varijable u dugoročnoj prognozi pacijenata lečenih putem pPCI nije razjašnjena. Cilj: Dokazati prediktivni značaj i odrediti najbolju vrednost Jung varijable u predviđanju jednogodišnjeg neželjenog kliničkog ishoda kod pacijenata sa STEMI lečenih pPCI, kao i dokazati njenu validnost na nezavisnoj populaciji. Metode: Sprovedena je prospektivna studija praćenja pacijenata sa STEMI lečenih pPCI tokom godinu dana; primarna studija je sprovedena u Institutu za kardiovaskularne bolesti Vojvodine (IKVBV), a validaciona u Vojnomedicinskoj akademiji (VMA). Ishodi studije: smrtni ishod, akutna srčana insuficijencija (AHF) i zbirni neželjeni ishod. Jung varijabla je računata prema formuli: sistolni krvni pritisak / (srčana frekvencija × godine života) × 100. Prediktivna vrednost Jung variable i prethodno etabliranih skorova rizika TIMI, PAMI i Zwolle je evaluiarana adekvatnim statističkim metodama. Rezultati: Od 647 pacijenata uključenih u primarnu studiju, umrlo je 70 (10.8%), dok je AHF imalo 42 (6.5%); od 418 pacijenata uključenih u validacionu studiju umrlo je 33 (7.9%), a 52 (12.4%) pacijenta je imalo AHF. U primarnoj studiji Jung variabla je bila prediktor smrtnog ishoda i zbirnog neželjenog ishoda, dok je u validacionoj studiji Jung varijabla bila nezavisni prediktor nastanka AHF (p<0.01). Za mortalitet, Jung variabla <= 2.35 je imala sensitivnost 74.3% i specifičnost 77.3%. U primarnoj studiji, C-statistike i 95% interval poverenja Jung varijable za jednogodišnji mortalitet i zbirni neželjeni ishod su bile dobre (0.784 (0.750-0.815) i 0.764 (0.729-0.796)) i poredive sa TIMI, PAMI i Zwolle skorovima (p>0.05). C-statistika za predikciju AHF i zbirnog neželjenog ishoda u validacionoj studiji je bila dobra (0.732 (0.655-0.809) i 0.721 (0.655-0.788)), ali manja u odnosu na ostale ispitivane skorove (p<0.01). Zaključak: Jung varijabla je nezavisni prediktor jednogodišnjeg smrtnog ishoda i zbirnog neželjenog ishoda pacijenata sa STEMI lečenih pPCI u IKVBV i nezavisni prediktor AHF u validacionoj studiji. Jung varijabla, TIMI, PAMI i ZWOLLE skorovi imaju dobar i porediv diskriminatorni kapacitet za sve praćene ishode u primarnoj studiji, dok je u validacionoj studiji Jung varijabla imala dobar diskriminatorni kapacitet za AHF i zbirni neželjeni ishod, ali manji u odnosu na ostale ispitivane skorove.
Background: Accurate eary risk stratification of patients with ST-elevation myocardial infarction (STEMI) is important in the management of this patients. Primary percutaneous coronary intervention (pPCI) in patients with STEMI has improved the outcome significantly and might have changed the relative contribution of different risk factors. Jung variable is a simple clinical risk index, designed to be used at initial presentation. It is predictive of in hospital mortality in STEMI patients treated with fibrinolysis, but it's long term predictive power in patients treated with pPCI is not elucidated. Aim: To determine the prognostic accuracy and cut off value of Jung variable for one year clinical outcomes in STEMI patients treated with pPCI, and to validate it in independent STEMI patients. Methods: Two prospective studies of consecutive patients treated with pPCI were conducted; primary study in Institute of Cardiovascular diseases of Vojvodina and validation study in Military medical academy. One-year clinical outcomes (all-cause mortality, acute heart failure (AHF) and combined end point) were assessed. The Jung variable was calculated using the equation: systolic blood/ (heart rate×age)×100. The predictive value of Jung variable and previously established scores TIMI, PAMI, and Zwolle were evaluated with adequate statistical analyses. Results: Out of 647 patients 70 (10.8%) died and 42 (6.5%) had AHF in primary, while out of 418 patients 33 (7.9%) died and 52 (12.4%) had AHF in validation study. In primary study Jung variable was independent predictor of mortality and combined end point, while it was independent predictor of AHF in validation study (p<0.01). Jung variable <= 2.35 had sensitivity 74.3% and specificity 77.3% for mortality. In a primary study, C-statistic and 95% confidence intervals of Jung variable for one-year mortality and for combined end point were well (0.784 (0.750-0.815) and 0.764 (0.729-0.796), respectively) and comparable to TIMI, PAMI and Zwolle risk score (p>0.05). C-statistic for predicting one-year AHF and combined clinical endpoint in a validation study was well (0.732 (0.655-0.809) and 0.721 (0.655-0.788), respectively), but lower than other risk scores (p<0.01). Conclusion: The Jung variable was independent predictor of one year mortality and combined end point in primary study STEMI patients treated with pPCI and independent predictor of AHF in validation study. The Jung variable, TIMI, PAMI, and Zwolle risk scores performed well and comparable for all clinical outcomes in primary, while in a validation study Jung variable performed well for AHF and combined end point, but significantly lower than other risk scores.

The aim of my diploma thesis are methods which are able to describe durability aspects of glued wooden joints. The main objective are methods which primarily includes hygrothermal stress acting in the glued joint of wooden adherend and determine the service life of the specific glued joint by modelling based on Arrhenius relation and the other methods. Other part of this work is aimed to shear tests of the samples exposed to different thermal and humidity conditions and for individual sets of exposures determining the glue penetration depth into the wooden adherend by EPI fluorescent analysis. These analytic methods are used for quality check of the glued joints, which are mostly based on strength, chemical or structural parameters of the tested sample.

The aim of this work is comparing three casting process simulation programs for porosity and microstructure prediction capabilities for die-casting. After confronting these results with experimentally measured data taken from real castings it could be said that simulation is pretty accurate for DAS microstructure prediction and hot-spot areas. Amount of measured porosity could not be compared with predicted values because specimens were not taken from the exact hot-spot areas.

Data mining is a new discipline that occurs with increasing amount of stored data and the increasing need to obtain the information hidden in them. It is focused on the mining of potentially useful information from large data sets and it lies at the intersection of statistics, machine learning, artificial intelligence, databases and other areas. The aim of this thesis is to present the process of data mining with an emphasis on its connection with statistics and to describe a selection of statistical methods widely used in this field and which were also used in the applied data mining problem in this thesis. Real data from purchases in the online store show that using different methods gives different results and interesting information about purchasing behavior, and also proves that not all methods are always applicable to all types of tasks.

The aim of this PhD thesis was an objective assessment of application of VNIR spectroscopy for predicting properties of forest soils. For each soil property were found the most appropriate combination of statistical methods for pre-processing (continuum removal, 1. derivation, 2. derivation) and processing (PLSR, PCR, SVM) of certain spectral bands. As generally successful shows a combination of methods 1. derivation and support vector machine throughout the VNIR spectral range (400-2500 nm). In some cases, however, they proved to other models. Among the best predictable features include pH, content of oxidizable carbon, aluminum, iron, silicon, or calcium (at higher concentrations). Not very high success rate prediction was found in indicators that take low values (sodium, manganese, aluminum or ferrous complexes). The results show that VNIR spectroscopy method is applicable for predicting properties of forest soils. It can not completely replace traditional analysis, but it can very well complement, especially in practice. For example, when the soil mapping can help thicken network data and refine the information better than other methods of spatial estimation. It is applicable in cases where it is required large amounts of data in a short timeframe and at minimal cost. It is suitable for monitoring trends over time, or for a quick survey of an area.

Náplní této práce je teoretický úvod a následné praktické zpracování tématu Molekulární signatura jako optimální multi-objektivní funkce s aplikací v predikci v onkogenomice. Úvodní kapitoly jsou zaměřeny na téma rakovina, zejména pak rakovina prsu a její podtyp triple negativní rakovinu prsu. Následuje literární přehled z oblasti optimalizačních metod, zejména se zaměřením na metaheuristické metody a problematiku strojového učení. Část se odkazuje na onkogenomiku a principy microarray a také na statistiku a s důrazem na výpočet p-hodnoty a bimodálního indexu. Praktická část je pak zaměřena na konkrétní průběh výzkumu a nalezené závěry, vedoucí k dalším krokům výzkumu. Implementace vybraných metod byla provedena v programech Matlab a R, s využitím dalších programovacích jazyků a to konkrétně programů Java a Python.

The purpose of this diploma thesis is comparing the quantity of porosity in die-cast aluminium casting with using simulation. The cast engine blocks were made by high pressure die-casting and a local squeeze casting (LSC) technology. There was also examined the relationship between porosity and mechanical properties of castings. There were evaluated and compared the results of my own measurements with data from previous students theses at the Faculty of Mechanical Engineering BUT. All data were tested by statistical tests. The comparison of simulation results and experimentally measured values was found a good programme-ability to simulate and estimate porosity.

Hra Fantasy Premier League poskytuje miliónom hráčov po celom svete možnosť stať sa na chvíľu manažérom svojho vlastného klubu. Výsledky a bodové ohodnotenie v hre závisia na správnom predvídaní, ako sa budú hráči chovať v skutočných futbalových zápasoch. Ak by pri tomto rozhodovaní pomáhal software na predikciu a analýzu budúcich výkonov hráčov, výsledky v hre sa môžu rapídne zlepšiť. Táto diplomová práca sa zaoberá návrhom a implementáciou predikčného modelu, ktorý využíva neurónové siete na predikcie časových radov počas celej sezóny v hre. Boli použité metódy na spracovanie dát o hráčoch a kluboch za posledné 4 sezóny. Výkonnosť a presnosť predikčných metód boli testované na dátach z poslednej sezóny Premier League a predikcie algoritmu sa vo väčšine prípadov blížili realite. Ak by sa užívateľ držal predikčného modelu v hre stopercentne, získal by väčší počet bodov ako bežný hráč, ktorý žiadny predikčný model nepoužíva.

This master thesis deals with issue of trade on commodity market, especially the gold. It uses the artificial intelligence resources, more accurate non-linear auregressive neural network. The purpose is the prediction of the gold prices by indicators which has impact on the gold.

This master thesis focuses on application of statistical methods in the processing of data. The first part of the thesis describes the theoretical foundations that are the basis for the practical part. Next part of this thesis describes the statistical and financial analysis and also design of an application that automate usage of statistical methods of regression analysis to predict the future economic situation development of the company. This thesis contains theory of time series methods and regression analysis.

This thesis is focus on monitoring state of machine parts that are under the most stress. Type of artificial intelligence used in this work is recurrent neural network and its modifications. Chosen type of neural network was used because of the sequential character of used data. This thesis is solving three problems. In first problem algorithm is trying to determine state of mill tool wear using recurrent neural network. Used method for monitoring state is indirect. Second Problem was focused on detecting fault of a bearing and classifying it to specific category. In third problem RNN is used to predict RUL of monitored bearing.

Knowledge discovery from databases is a complex issue involving integration, data preparation, data mining using machine learning methods and visualization of results. The thesis deals with the whole process of knowledge discovery, especially with the issue of data warehousing, where it offers the design and implementation of a specific data warehouse for the company ROI Hunter, a.s. In the field of data mining, the work focuses on the classification and forecasting of the advertising data available from the prepared data warehouse and, in particular, on the decision tree classification. When predicting the development of new ads, emphasis is put on the rationale for the prediction as well as the proposal to adjust the ad settings so that the prediction ends positively and, with a certain likelihood, the ads actually get better results.

The goal of the thesis was to study the indoor air quality measurement focusing on the concentration of carbon dioxide. Within the theoretical part, data mining including basic classification methods and approaches to dimensionality reduction was introduced. In addition, the principles of the developed system within IoTCloud project and available possibilities for measurement of necessary quantities were studied. In the practical part, the suitable sensors for given rooms were selected and long-term measurement was performed. Measured data was used to create the system for window opening detection and for the design of appropriate way of air change regulation in a room. The aim of regulation was to improve air quality using natural ventilation.

This diploma thesis deals with the analysis of volume changes of selected silicate based composites. The first part introduces the issue of volume changes of concrete, especially the shrinkage process, two selected models for shrinkage prediction, describes configuration and course of performed tests, as well as the processing of measured experimental data. The second part contains input data modified by the program GTDiPS and another calculated parameters. The result of the work is the evaluation of the influence of thermal dilation on the development of shrinkage of young concrete. Another output is the comparison of accuracy of the two selected prediction models to capture the course of measured shrinkage.

The goal of this diploma thesis is to design and develop a system for collecting, processing and storing data from large IoT networks. The developed system introduces a complex solution able to process data from various IoT networks using Apache Hadoop ecosystem. The data are real-time processed and stored in a NoSQL database, but the data are also stored  in the file system for a potential later processing. The system is optimized and tested using data from IQRF network. The data stored in the NoSQL database are visualized and the system periodically generates derived predictions. Users are connected to this system via an information system, which is able to automatically generate notifications when monitored values are out of range.

Diplomová práce pojednává o stabilizaci diskrétního logistického modelu pomocí několika řídících metod. Je zde provedena především stabilizace rovnováh, 2-periodických cyklů a 3-periodických cyklů. Ke stabilizaci systému je využito proporčního zpětně-vazebního řízení, zpětně-vazebního řízení s časovým zpožděním a řízení založeného na predikci. U každé metody je diskutovaná stabilizační množina pro řídící zesilovač spolu s oblastmi stability pro odpovídající kontrolovaná řešení. Všechny teoretické výsledky jsou ilustrovány grafickými interpretacemi v softwaru MATLAB. Podpůrné výpočty jsou provedeny pomocí softwaru Maple.

This diploma thesis main goal is to predict primary bankruptcy indicators in construction sector based on model. This model is built by advanced classifier methods such as classification trees and logistic regression. Applicability of prediction depends on quality of built model and therefore verification of classification accuracy is essential for bankruptcy prediction. The theoretical part is focused on construction sector in the EU specializing on basic financial analysis including ratio indicators and frequently used models for predicting the bankruptcy of enterprises. The practical part consists of model creation and evaluation for each observed time period. The end of thesis is dealing with model evaluation based on given criteria.

The thesis analyzes the accuracy of the multi-step inflation forecast using se-lected methods of machine learning through inflationary factors in the Visegrad group countries. The methods that were applied in the work analysis are the re-gression of tree methods and the algorithm method to the k-nearest neighbors. Based on the regression tree method, we are able to identify factors that are most prominent in price level development. The output of the analysis consists of 8 models, the suitability and accuracy of which are discussed. The results of the em-pirical analysis are compared with the assumptions that were presented before the analysis has begun. This suggests that methods are not suitable for multi-step inflation prediction.

This thesis deals with evaluating accuracy of GDP prediction. Several indicators will be calculated to determine the accuracy, e.g. mean absolute error, root mean squared error, average forecast error and Theil coefficient. First section presents history of economic theory understanding, general description of DSGE model and description of national models. Countries with most and least reliable prediction models are identified through analysis. GEAR model of German national bank ended up as most reliable, while ÉIRE model of National Bank of Ireland is on the other side with lowest score. If company is seeking the most reliable information about future changes of GDP, it shall look for predictions of German national bank. Having more precise predictions leads to sustainable and stable business sphere.

Title: Methods of artificial intelligence and their use in prediction Author: Lubomír Šerý Department: Department of Probability and Mathematical Statistics Supervisor: Ing. Marek Omelka, Ph.D., Department of Probability and Mathe- matical Statistics Abstract: In the presented thesis we study field of artificial intelligence, in par- ticular we study part dedicated to artificial neural networks. At the beginning, concept of artificial neural networks is introduced and compared to it's biological base. Afterwards, we also compare neural networks to some generalized linear models. One of the main problems of neural networks is their learning. Therefore biggest part of this work is dedicated to learning algorithms, especially to pa- rameter estimation and specific computational aspects. In this part we attempt to bring in an overview of internal structure of neural network and to propose enhancement of learning algorithm. There are lots of techniques for enhancing and enriching basic model of neural networks. Some of these improvements are, together with genetic algorithms, introduced at the end of this work. At the very end of this work simulations are presented, where we attempt to verify some of the introduced theoretical assumptions and conclusions. Main simulation is an application of concept of neural...

Charles University, Faculty of Pharmacy in Hradec Králové Department of Pharmacology & Toxicology Student: Helena Zálešáková Supervisor: PharmDr. Marie Vopršálová, CSc. Title of diploma thesis: In vitro Methods for the Prediction of Blood Brain Barrier Penetration This thesis deals with the correlation between two in vitro models simulating the blood- brain barrier (HEB, hematoencephalic barrier) and their comparison in terms of practical use. These are the PAMPA (Parallel Artificial Membrane Permeability Assay) method and the MDCK (Madin-Darby Canine Kidney) cell line, which are models for potential central nervous system (CNS) penetration screening. Within this work, a set of sixteen standard drugs were measured. The procedure was similar in both methods in order to obtain information on the amount of test substances passing through the membrane from the donor portion of the plate to the acceptor. The concentration in the donor portion was measured by UV-VIS spectrophotometry. The main difference between these methods is the membrane through which the substances penetrate. In the case of PAMPA, a lipid solution that has been isolated from pig brain (PBL, polar brain lipid) is used. This lipid simulates the phospholipid membrane of the brain capillary endothelium. In the MDCK model, the membrane...

In this diploma thesis we tried to model liquidity risk with the use of copula functions. We presented theoretical background for both areas, theory of copulas and liquidity risk. Empirical work is focusing more on the liquidity in CEE countries and especially in Slovakia and Czech Republic. In this part we created a market liquidity index for Slovak financial market and analyzed affects of liquidity problems during the financial crisis on the overall liquidity in this market. Furthermore, we modelled the possibility of market contagion of liquidity risk in Czech banking sector, based on loan-to-deposit ratios of Komerční banka, Česká spořitelna and GE Money Bank.

In this thesis we will further develop the algorithm of homologous prediction of tertiary RNA structure. The algorithm was originally created and implemented in my bachelor thesis. We will focus on further automatization of algorithm implementation and we are going to make it easier to use. The user will be able to predict tertiary structure of RNA based only on target structure sequence. The algorithm will be also extended to use multiple template structures for prediction and it will be able to firstly predict the secondary structure of the target molecule. Both of these modifications should lead to more precise prediction by restricting the search space and reducing the size of unconserved regions of the predicted structure.

Nowadays, electricity production from photovoltaics power plants is becoming important increasingly. In order to set production to other power plants, it is necessary to predict the generation of electricity from these sources. The thesis is mainly devoted to models for short-term prediction, which is based on weather forecast. The models were designated by beta regression and linear regression with transformed explanatory variable. One part of thesis is devoted to Clear sky model, which is used to estimated the maximum possible production at given hour. 1

This diploma thesis deals with the creation and testing of trading strategies based on technical analysis and prediction of time series using neural networks. The the-oretical part introduces the reader to basic methods of market analysis, especially with technical analysis. In the practical part there are analyzed available options for testing trading strategies and designed functions and structure, which are used to implement web application.