Bayesian methods for estimation, optimization and experimental design

Mathematical models often contain parameters that need to be calibrated from measured data. The emergence of efficient Markov Chain Monte Carlo (MCMC) methods has made the Bayesian approach a standard tool in quantifying the uncertainty in the parameters. With MCMC, the parameter estimation problem can be solved in a fully statistical manner, and the whole distribution of the parameters can be explored, instead of obtaining point estimates and using, e.g., Gaussian approximations. In this thesis, MCMC methods are applied to parameter estimation problems in chemical reaction engineering, population ecology, and climate modeling. Motivated by the climate model experiments, the methods are developed further to make them more suitable for problems where the model is computationally intensive. After the parameters are estimated, one can start to use the model for various tasks. Two such tasks are studied in this thesis: optimal design of experiments, where the task is to design the next measurements so that the parameter uncertainty is minimized, and model-based optimization, where a model-based quantity, such as the product yield in a chemical reaction model, is optimized. In this thesis, novel ways to perform these tasks are developed, based on the output of MCMC parameter estimation. A separate topic is dynamical state estimation, where the task is to estimate the dynamically changing model state, instead of static parameters. For example, in numerical weather prediction, an estimate of the state of the atmosphere must constantly be updated based on the recently obtained measurements. In this thesis, a novel hybrid state estimation method is developed, which combines elements from deterministic and random sampling methods.

MCMC Analysis for Optimization of Stochastic Models

In any decision making under uncertainties, the goal is mostly to minimize the expected cost. The minimization of cost under uncertainties is usually done by optimization. For simple models, the optimization can easily be done using deterministic methods.However, many models practically contain some complex and varying parameters that can not easily be taken into account using usual deterministic methods of optimization. Thus, it is very important to look for other methods that can be used to get insight into such models. MCMC method is one of the practical methods that can be used for optimization of stochastic models under uncertainty. This method is based on simulation that provides a general methodology which can be applied in nonlinear and non-Gaussian state models. MCMC method is very important for practical applications because it is a uni ed estimation procedure which simultaneously estimates both parameters and state variables. MCMC computes the distribution of the state variables and parameters of the given data measurements. MCMC method is faster in terms of computing time when compared to other optimization methods. This thesis discusses the use of Markov chain Monte Carlo (MCMC) methods for optimization of Stochastic models under uncertainties .The thesis begins with a short discussion about Bayesian Inference, MCMC and Stochastic optimization methods. Then an example is given of how MCMC can be applied for maximizing production at a minimum cost in a chemical reaction process. It is observed that this method performs better in optimizing the given cost function with a very high certainty.

Viestitaktiikan kehittyminen yhtymän viestijärjestelmän käyttöönoton jälkeen

Tutkimus käynnistyi Maanpuolustuskorkeakoulun taktiikan laitoksen esittäessä aihepiiriä tutkittavaksi. Tutkimuksen tavoitteena on ollut lisätä tietämystä viestitaktiikan kehittymisestä yhtymän viestijärjestelmän käyttöönoton jälkeen 1980 – 2000-luvuilla osana operatiivistaktisten toimintaperiaatteiden ja -tapojen kehittymistä. Tutkimuksella on pyritty syventämään tietämystä taktisten periaatteiden muutoksista viestitaktiikan näkökulmasta. Tutkimuksessa tarkasteltiin maavoimien YVI-järjestelmillä varustettujen yhtymien viestitaktiikkaa sekä niissä tapahtuneita muutoksia. Muutoksia tarkasteltaessa tutkimuksessa keskityttiin käsitykseen viestitaktiikasta, viestitaktisiin periaatteisiin sekä viestipäällikköön ja hänen toimintakenttäänsä. Viestitaktisia periaatteita ja niissä tapahtuneita muutoksia vertailtiin myös yleisiin taktisiin periaatteisiin ja niiden painotuksissa tapahtuneisiin muutoksiin. Tutkimus on luonteeltaan kvalitatiivinen. Tutkimusongelmia lähestyttiin fenomenografisella tutkimusotteella, jossa tavoitteena on kuvailla, analysoida ja ymmärtää erilaisia käsityksiä ilmiöistä sekä käsitysten keskinäisistä suhteista. Lähdeaineiston muodostivat 18 viestitaktiikan asiantuntijan kokemusperäiset käsitykset viestitaktiikasta ja sen kehittymisestä YVIjärjestelmien käyttöönoton jälkeen. Käsityksistä muodostettiin merkitys- ja kuvauskategorioiden sekä tutkijan esiymmärryksen pohjalta induktiivisen päättelyn avulla tutkimuksen varsinaiset johtopäätökset. Tutkimushenkilöiden käsitysten sekä taktiikan ja viestitaktiikan aikaisempien määritelmien perusteella johtopäätöksenä määritettiin, että viestitaktiikka on tehtävän toteuttamiseen käytettävissä olevan viestillisen kapasiteetin optimaalista suunnittelua, soveltamista ja käyttöä viestivoimana haluttujen päämäärien saavuttamiseksi ja viestitaisteluiden voittamiseksi. Viestitaktikointi edellyttää viestitaisteluun liittyvien keinojen tuntemista sekä taitoa soveltaa niitä käytännössä. Tutkimustulosten perusteella keskeisiksi viestitaktisiksi periaatteiksi tärkeysjärjestyksessä muodostuivat - päämäärän ja tehtävän selkeys - varautuminen odottamattomiin tilanteen vaihteluihin - yksinkertaisuus - aktiivisuus ja oma-aloitteisuus. Keskeisiksi merkitystään lisänneiksi viestitaktisiksi periaatteiksi muodostuivat - voimien vaikutuksen keskittäminen - joukkojen ja voimien jakaminen (reservi) - varautuminen odottamattomiin tilanteen vaihteluihin - salaaminen ja harhauttaminen - turvallisuus. Selkeimpänä viestipäällikön tehtävien muutoksena pidettiin siirtymistä yksityiskohtaisesta viestiyhteyksien suunnittelijasta kokonaisvaltaiseksi yhtymän viestitoiminnan johtajaksi. Tutkimustulosten ja aikaisempien määritelmien perusteella johtopäätöksenä määritettiin, että viestipäällikkö johtaa yhtymän viestitoimintaa komentajan antamien vaatimusten mukaisesti ja vastaa yhtymän johtoryhmän jäsenenä viestitaktisista ratkaisuista haluttujen päämäärien saavuttamiseksi ja viestitaisteluiden voittamiseksi. Viestipäälliköltä edellytetään viestitaisteluun liittyvien keinojen tuntemista sekä taitoa soveltaa niitä käytännössä. Tutkimuksen mukaan yhtymän viestitaktiikkaan merkittävimmin vaikuttaneita tekijöitä olivat yhtymän viestijärjestelmien käyttöönotto, uusien esikunta- ja viestiyksiköiden kehittäminen, kiinteän viestiverkon ja johtamisjärjestelmäalan merkityksen kasvaminen, käytettävien tekniikoiden kehittyminen sekä joukkojen ja johtoportaiden tiedonsiirtotarpeiden kasvaminen. Viestitaktiikan osalta voidaan todeta deterministisen näkemyksen taistelusta ja taistelutilasta muuttuneen yleisten taktisten periaatteiden muutosten mukaisesti aikaisempaa monimuotoisempaan ja rohkeampaan, voluntaarisempaan, suuntaan.

Voimalaitoksen käytönaikaisen optimoinnin määrittely

Tässä diplomityössä määritellään biopolttoainetta käyttävän voimalaitoksen käytönaikainen tuotannon optimointimenetelmä. Määrittelytyö liittyy MW Powerin MultiPower CHP –voimalaitoskonseptin jatkokehitysprojektiin. Erilaisten olemassa olevien optimointitapojen joukosta valitaan tarkoitukseen sopiva, laitosmalliin ja kustannusfunktioon perustuva menetelmä, jonka tulokset viedään automaatiojärjestelmään PID-säätimien asetusarvojen muodossa. Prosessin mittaustulosten avulla lasketaan laitoksen energia- ja massataseet, joiden tuloksia käytetään seuraavan optimointihetken lähtötietoina. Optimoinnin kohdefunktio on kustannusfunktio, jonka termit ovat voimalaitoksen käytöstä aiheutuvia tuottoja ja kustannuksia. Prosessia optimoidaan säätimille annetut raja-arvot huomioiden niin, että kokonaiskate maksimoituu. Kun laitokselle kertyy käyttöikää ja historiadataa, voidaan prosessin optimointia nopeuttaa hakemalla tilastollisesti historiadatasta nykytilanteen olosuhteita vastaava hetki. Kyseisen historian hetken katetta verrataan kustannusfunktion optimoinnista saatuun katteeseen. Paremman katteen antavan menetelmän laskemat asetusarvot otetaan käyttöön prosessin ohjausta varten. Mikäli kustannusfunktion laskenta eikä historiadatan perusteella tehty haku anna paranevaa katetta, niiden laskemia asetusarvoja ei oteta käyttöön. Sen sijaan optimia aletaan hakea deterministisellä optimointialgoritmilla, joka hakee nykyhetken ympäristöstä paremman katteen antavia säätimien asetusarvoja. Säätöjärjestelmä on mahdollista toteuttaa myös tulevaisuutta ennustavana. Työn käytännön osuudessa voimalaitosmalli luodaan kahden eri mallinnusohjelman avulla, joista toisella kuvataan kattilan ja toisella voimalaitosprosessin toimintaa. Mallinnuksen tuloksena saatuja prosessiarvoja hyödynnetään lähtötietoina käyttökatteen laskennassa. Kate lasketaan kustannusfunktion perusteella. Tuotoista suurimmat liittyvät sähkön ja lämmön myyntiin sekä tuotantotukeen, ja suurimmat kustannukset liittyvät investoinnin takaisinmaksuun ja polttoaineen ostoon. Kustannusfunktiolle tehdään herkkyystarkastelu, jossa seurataan katteen muutosta prosessin teknisiä arvoja muutettaessa. Tuloksia vertaillaan referenssivoimalaitoksella suoritettujen verifiointimittausten tuloksiin, ja havaitaan, että tulokset eivät ole täysin yhteneviä. Erot johtuvat sekä mallinnuksen puutteista että mittausten lyhyehköistä tarkasteluajoista. Automatisoidun optimointijärjestelmän käytännön toteutusta alustetaan määrittelemällä käyttöön otettava optimointitapa, siihen liittyvät säätöpiirit ja tarvittavat lähtötiedot. Projektia tullaan jatkamaan järjestelmän ohjelmoinnilla, testauksella ja virityksellä todellisessa voimalaitosympäristössä ja myöhemmin ennustavan säädön toteuttamisella.

Differential Evolution approach and parameter estimation of chaotic

Parameter estimation still remains a challenge in many important applications. There is a need to develop methods that utilize achievements in modern computational systems with growing capabilities. Owing to this fact different kinds of Evolutionary Algorithms are becoming an especially perspective field of research. The main aim of this thesis is to explore theoretical aspects of a specific type of Evolutionary Algorithms class, the Differential Evolution (DE) method, and implement this algorithm as codes capable to solve a large range of problems. Matlab, a numerical computing environment provided by MathWorks inc., has been utilized for this purpose. Our implementation empirically demonstrates the benefits of a stochastic optimizers with respect to deterministic optimizers in case of stochastic and chaotic problems. Furthermore, the advanced features of Differential Evolution are discussed as well as taken into account in the Matlab realization. Test "toycase" examples are presented in order to show advantages and disadvantages caused by additional aspects involved in extensions of the basic algorithm. Another aim of this paper is to apply the DE approach to the parameter estimation problem of the system exhibiting chaotic behavior, where the well-known Lorenz system with specific set of parameter values is taken as an example. Finally, the DE approach for estimation of chaotic dynamics is compared to the Ensemble prediction and parameter estimation system (EPPES) approach which was recently proposed as a possible solution for similar problems.

Epävarmuustarkastelun soveltaminen deterministisissä turvallisuusanalyyseissä

Tässä diplomityössä esitetään selvitys käytössä olevista deterministisistä turvallisuusanalyysimenetelmistä. Deterministisillä turvallisuusanalyyseillä arvioidaan ydinvoimalaitosten turvallisuutta eri käyttötilojen aikana. Voimalaitoksen turvallisuusjärjestelmät mitoitetaan deterministisen turvallisuusanalyysin tulosten perusteella. Deterministiset turvallisuusanalyysit voidaan laatia konservatiivista tai tilastollista menetelmää käyttäen. Konservatiivinen menetelmä pyrkii mallintamaan tarkasteltavan tilanteen siten, että laitoksen todellinen käyttäytyminen on hyvällä varmuudella lievempää kuin analyysitulos. Konservatiivisessa menetelmässä analyysin epävarmuudet huomioidaan konservatiivisilla oletuksilla. Tilastollinen menetelmä perustuu parhaan arvion menetelmään eli pyrkimykseen mallintaa laitoksen käyttäytyminen mahdollisimman todenmukaisesti. Tilastollisessa menetelmässä analyysin epävarmuudet määritetään systemaattisesti tilastomatematiikan keinoin. Työssä painotetaan tilastollisen analyysin epävarmuuksien määritykseen käytettäviä epävarmuustarkastelumenetelmiä. Diplomityön laskennallisessa osassa vertaillaan deterministisen turvallisuusanalyysin laadintaan käytettäviä menetelmiä termohydraulisen turvallisuusanalyysiesimerkin laskennan kautta. Laskennassa tarkasteltavana onnettomuutena on Olkiluoto 3-laitosyksikössä tapahtuva primäärijäähdytepiirin putkikatkosta aiheutuva jäähdytteenmenetysonnettomuus. Lasketun esimerkkitapauksen perusteella tilastollista ja konservatiivista menetelmää voidaan pitää vaihtoehtoisina turvallisuusanalyysin laadintaan. Molemmat analyysit tuottivat hyväksyttäviä ja toisilleen verrannollisia tuloksia, joiden suuruusluokka on sama.

Numbers and Languages

The thesis presents results obtained during the authors PhD-studies. First systems of language equations of a simple form consisting of just two equations are proved to be computationally universal. These are systems over unary alphabet, that are seen as systems of equations over natural numbers. The systems contain only an equation X+A=B and an equation X+X+C=X+X+D, where A, B, C and D are eventually periodic constants. It is proved that for every recursive set S there exists natural numbers p and d, and eventually periodic sets A, B, C and D such that a number n is in S if and only if np+d is in the unique solution of the abovementioned system of two equations, so all recursive sets can be represented in an encoded form. It is also proved that all recursive sets cannot be represented as they are, so the encoding is really needed. Furthermore, it is proved that the family of languages generated by Boolean grammars is closed under injective gsm-mappings and inverse gsm-mappings. The arguments apply also for the families of unambiguous Boolean languages, conjunctive languages and unambiguous languages. Finally, characterizations for morphisims preserving subfamilies of context-free languages are presented. It is shown that the families of deterministic and LL context-free languages are closed under codes if and only if they are of bounded deciphering delay. These families are also closed under non-codes, if they map every letter into a submonoid generated by a single word. The family of unambiguous context-free languages is closed under all codes and under the same non-codes as the families of deterministic and LL context-free languages.

Computational chemistry studies of wood-derived lignans

This thesis is based on computational chemistry studies on lignans, focusing on the naturally occurring lignan hydroxymatairesinol (HMR) (Papers I II) and on TADDOL-like conidendrin-based chiral 1,4-diol ligands (LIGNOLs) (Papers III V). A complete quantum chemical conformational analysis on HMR was previously conducted by Dr. Antti Taskinen. In the works reported in this thesis, HMR was further studied by classical molecular dynamics (MD) simulations in aqueous solution including torsional angle analysis, quantum chemical solvation e ect study by the COnductorlike Screening MOdel (COSMO), and hydrogen bond analysis (Paper I), as well as from a catalytic point of view including protonation and deprotonation studies at di erent levels of theory (Paper II). The computational LIGNOL studies in this thesis constitute a multi-level deterministic structural optimization of the following molecules: 1,1-diphenyl (2Ph), two diastereomers of 1,1,4-triphenyl (3PhR, 3PhS), 1,1,4,4-tetraphenyl (4Ph) and 1,1,4,4-tetramethyl (4Met) 1,4-diol (Paper IV) and a conformational solvation study applying MD and COSMO (Paper V). Furthermore, a computational study on hemiketals in connection with problems in the experimental work by Docent Patrik Eklund's group synthesizing the LIGNOLs based on natural products starting from HMR, is shortly described (Paper III).

Tooth-Coil Permanent Magnet Synchronous Machine Design for Special Applications

This doctoral thesis presents a study on the design of tooth-coil permanent magnet synchronous machines. The electromagnetic properties of concentrated non-overlapping winding permanent magnet synchronous machines, or simply tooth-coil permanent magnet synchronous machines (TC-PMSMs), are studied in details. It is shown that current linkage harmonics play the deterministic role in the behavior of this type of machines. Important contributions are presented as regards of calculation of parameters of TC-PMSMs,particularly the estimation of inductances. The current linkage harmonics essentially define the air-gap harmonic leakage inductance, rotor losses and localized temporal inductance variation. It is proven by FEM analysis that inductance variation caused by the local temporal harmonic saturation results in considerable torque ripple, and can influence on sensorless control capabilities. Example case studies an integrated application of TC-IPMSMs in hybrid off-highway working vehicles. A methodology for increasing the efficiency of working vehicles is introduced. It comprises several approaches – hybridization, working operations optimization, component optimization and integration. As a result of component optimization and integration, a novel integrated electro-hydraulic energy converter (IEHEC) for off-highway working vehicles is designed. The IEHEC can considerably increase the operational efficiency of a hybrid working vehicle. The energy converter consists of an axial-piston hydraulic machine and an integrated TCIPMSM being built on the same shaft. The compact assembly of the electrical and hydraulic machines enhances the ability to find applications for such a device in the mobile environment of working vehicles.Usage of hydraulic fluid, typically used in working actuators, enables direct-immersion oil cooling of designed electrical machine, and further increases the torque- and power- densities of the whole device.

Stochastic approximation methods in stochastic optimization

Stochastic approximation methods for stochastic optimization are considered. Reviewed the main methods of stochastic approximation: stochastic quasi-gradient algorithm, Kiefer-Wolfowitz algorithm and adaptive rules for them, simultaneous perturbation stochastic approximation (SPSA) algorithm. Suggested the model and the solution of the retailer's profit optimization problem and considered an application of the SQG-algorithm for the optimization problems with objective functions given in the form of ordinary differential equation.

The succession dynamics of a macroalgal community after a flood disturbance in a tropical stream from São Paulo State, southeastern Brazil

The succession dynamics of a macroalgal community in a tropical stream (20º58' S and 49º25' W) was investigated after disturbance by a sequence of intensive rains. High precipitation levels caused almost complete loss of the macroalgal community attached to the substratum and provided a strong pressure against its immediate re-establishment. After this disturbance, a weekly sampling program from May 1999 to January 2000 was established to investigate macroalgal recolonization. The community changed greatly throughout the succession process. The number of species varied from one to seven per sampling. Global abundance of macroalgal community did not reveal a consistent temporal pattern of variation. In early succession stages, the morphological form of tufts dominated, followed by unbranched filaments. Latter succession stages showed the almost exclusive occurrence of gelatinous forms, including filaments and colonies. The succession trajectory was mediated by phosphorus availability in which community composition followed a scheme of changes in growth forms. However, we believe that deterministic and stochastic processes occur in lotic ecosystems, but they are dependent on the length of time considered in the succession analyses.

Mathematical modeling and optimal control of malaria

Malaria continues to infect millions and kill hundreds of thousands of people worldwide each year, despite over a century of research and attempts to control and eliminate this infectious disease. Challenges such as the development and spread of drug resistant malaria parasites, insecticide resistance to mosquitoes, climate change, the presence of individuals with subpatent malaria infections which normally are asymptomatic and behavioral plasticity in the mosquito hinder the prospects of malaria control and elimination. In this thesis, mathematical models of malaria transmission and control that address the role of drug resistance, immunity, iron supplementation and anemia, immigration and visitation, and the presence of asymptomatic carriers in malaria transmission are developed. A within-host mathematical model of severe Plasmodium falciparum malaria is also developed. First, a deterministic mathematical model for transmission of antimalarial drug resistance parasites with superinfection is developed and analyzed. The possibility of increase in the risk of superinfection due to iron supplementation and fortification in malaria endemic areas is discussed. The model results calls upon stakeholders to weigh the pros and cons of iron supplementation to individuals living in malaria endemic regions. Second, a deterministic model of transmission of drug resistant malaria parasites, including the inflow of infective immigrants, is presented and analyzed. The optimal control theory is applied to this model to study the impact of various malaria and vector control strategies, such as screening of immigrants, treatment of drug-sensitive infections, treatment of drug-resistant infections, and the use of insecticide-treated bed nets and indoor spraying of mosquitoes. The results of the model emphasize the importance of using a combination of all four controls tools for effective malaria intervention. Next, a two-age-class mathematical model for malaria transmission with asymptomatic carriers is developed and analyzed. In development of this model, four possible control measures are analyzed: the use of long-lasting treated mosquito nets, indoor residual spraying, screening and treatment of symptomatic, and screening and treatment of asymptomatic individuals. The numerical results show that a disease-free equilibrium can be attained if all four control measures are used. A common pitfall for most epidemiological models is the absence of real data; model-based conclusions have to be drawn based on uncertain parameter values. In this thesis, an approach to study the robustness of optimal control solutions under such parameter uncertainty is presented. Numerical analysis of the optimal control problem in the presence of parameter uncertainty demonstrate the robustness of the optimal control approach that: when a comprehensive control strategy is used the main conclusions of the optimal control remain unchanged, even if inevitable variability remains in the control profiles. The results provide a promising framework for the design of cost-effective strategies for disease control with multiple interventions, even under considerable uncertainty of model parameters. Finally, a separate work modeling the within-host Plasmodium falciparum infection in humans is presented. The developed model allows re-infection of already-infected red blood cells. The model hypothesizes that in severe malaria due to parasite quest for survival and rapid multiplication, the Plasmodium falciparum can be absorbed in the already-infected red blood cells which accelerates the rupture rate and consequently cause anemia. Analysis of the model and parameter identifiability using Markov chain Monte Carlo methods is presented.

Semantic and fuzzy modelling for human behaviour recognition in smart spaces : a case study on ambient assisted living.

Human activity recognition in everyday environments is a critical, but challenging task in Ambient Intelligence applications to achieve proper Ambient Assisted Living, and key challenges still remain to be dealt with to realize robust methods. One of the major limitations of the Ambient Intelligence systems today is the lack of semantic models of those activities on the environment, so that the system can recognize the speci c activity being performed by the user(s) and act accordingly. In this context, this thesis addresses the general problem of knowledge representation in Smart Spaces. The main objective is to develop knowledge-based models, equipped with semantics to learn, infer and monitor human behaviours in Smart Spaces. Moreover, it is easy to recognize that some aspects of this problem have a high degree of uncertainty, and therefore, the developed models must be equipped with mechanisms to manage this type of information. A fuzzy ontology and a semantic hybrid system are presented to allow modelling and recognition of a set of complex real-life scenarios where vagueness and uncertainty are inherent to the human nature of the users that perform it. The handling of uncertain, incomplete and vague data (i.e., missing sensor readings and activity execution variations, since human behaviour is non-deterministic) is approached for the rst time through a fuzzy ontology validated on real-time settings within a hybrid data-driven and knowledgebased architecture. The semantics of activities, sub-activities and real-time object interaction are taken into consideration. The proposed framework consists of two main modules: the low-level sub-activity recognizer and the high-level activity recognizer. The rst module detects sub-activities (i.e., actions or basic activities) that take input data directly from a depth sensor (Kinect). The main contribution of this thesis tackles the second component of the hybrid system, which lays on top of the previous one, in a superior level of abstraction, and acquires the input data from the rst module's output, and executes ontological inference to provide users, activities and their in uence in the environment, with semantics. This component is thus knowledge-based, and a fuzzy ontology was designed to model the high-level activities. Since activity recognition requires context-awareness and the ability to discriminate among activities in di erent environments, the semantic framework allows for modelling common-sense knowledge in the form of a rule-based system that supports expressions close to natural language in the form of fuzzy linguistic labels. The framework advantages have been evaluated with a challenging and new public dataset, CAD-120, achieving an accuracy of 90.1% and 91.1% respectively for low and high-level activities. This entails an improvement over both, entirely data-driven approaches, and merely ontology-based approaches. As an added value, for the system to be su ciently simple and exible to be managed by non-expert users, and thus, facilitate the transfer of research to industry, a development framework composed by a programming toolbox, a hybrid crisp and fuzzy architecture, and graphical models to represent and con gure human behaviour in Smart Spaces, were developed in order to provide the framework with more usability in the nal application. As a result, human behaviour recognition can help assisting people with special needs such as in healthcare, independent elderly living, in remote rehabilitation monitoring, industrial process guideline control, and many other cases. This thesis shows use cases in these areas.

Deep generative models for facial keypoints detection

A new area of machine learning research called deep learning, has moved machine learning closer to one of its original goals: artificial intelligence and general learning algorithm. The key idea is to pretrain models in completely unsupervised way and finally they can be fine-tuned for the task at hand using supervised learning. In this thesis, a general introduction to deep learning models and algorithms are given and these methods are applied to facial keypoints detection. The task is to predict the positions of 15 keypoints on grayscale face images. Each predicted keypoint is specified by an (x,y) real-valued pair in the space of pixel indices. In experiments, we pretrained deep belief networks (DBN) and finally performed a discriminative fine-tuning. We varied the depth and size of an architecture. We tested both deterministic and sampled hidden activations and the effect of additional unlabeled data on pretraining. The experimental results show that our model provides better results than publicly available benchmarks for the dataset.

Defining Contexts in Context-Free Grammars

This thesis introduces an extension of Chomsky’s context-free grammars equipped with operators for referring to left and right contexts of strings.The new model is called grammar with contexts. The semantics of these grammars are given in two equivalent ways — by language equations and by logical deduction, where a grammar is understood as a logic for the recursive definition of syntax. The motivation for grammars with contexts comes from an extensive example that completely defines the syntax and static semantics of a simple typed programming language. Grammars with contexts maintain most important practical properties of context-free grammars, including a variant of the Chomsky normal form. For grammars with one-sided contexts (that is, either left or right), there is a cubic-time tabular parsing algorithm, applicable to an arbitrary grammar. The time complexity of this algorithm can be improved to quadratic,provided that the grammar is unambiguous, that is, it only allows one parsefor every string it defines. A tabular parsing algorithm for grammars withtwo-sided contexts has fourth power time complexity. For these grammarsthere is a recognition algorithm that uses a linear amount of space. For certain subclasses of grammars with contexts there are low-degree polynomial parsing algorithms. One of them is an extension of the classical recursive descent for context-free grammars; the version for grammars with contexts still works in linear time like its prototype. Another algorithm, with time complexity varying from linear to cubic depending on the particular grammar, adapts deterministic LR parsing to the new model. If all context operators in a grammar define regular languages, then such a grammar can be transformed to an equivalent grammar without context operators at all. This allows one to represent the syntax of languages in a more succinct way by utilizing context specifications. Linear grammars with contexts turned out to be non-trivial already over a one-letter alphabet. This fact leads to some undecidability results for this family of grammars