Modeling a Network of Heat Exchangers

This research work addresses the problem of building a mathematical model for the given system of heat exchangers and to determine the temperatures, pressures and velocities at the intermediate positions. Such model could be used in nding an optimal design for such a superstructure. To limit the size and computing time a reduced network model was used. The method can be generalized to larger network structures. A mathematical model which includes a system of non-linear equations has been built and solved according to the Newton-Raphson algorithm. The results obtained by the proposed mathematical model were compared with the results obtained by the Paterson approximation and Chen's Approximation. Results of this research work in collaboration with a current ongoing research at the department will optimize the valve positions and hence, minimize the pumping cost and maximize the heat transfer of the system of heat exchangers.

Feasibility of industrial implementation of laser cutting into paper making machine

Laser cutting implementation possibilities into paper making machine was studied as the main objective of the work. Laser cutting technology application was considered as a replacement tool for conventional cutting methods used in paper making machines for longitudinal cutting such as edge trimming at different paper making process and tambour roll slitting. Laser cutting of paper was tested in 70’s for the first time. Since then, laser cutting and processing has been applied for paper materials with different level of success in industry. Laser cutting can be employed for longitudinal cutting of paper web in machine direction. The most common conventional cutting methods include water jet cutting and rotating slitting blades applied in paper making machines. Cutting with CO2 laser fulfils basic requirements for cutting quality, applicability to material and cutting speeds in all locations where longitudinal cutting is needed. Literature review provided description of advantages, disadvantages and challenges of laser technology when it was applied for cutting of paper material with particular attention to cutting of moving paper web. Based on studied laser cutting capabilities and problem definition of conventional cutting technologies, preliminary selection of the most promising application area was carried out. Laser cutting (trimming) of paper web edges in wet end was estimated to be the most promising area where it can be implemented. This assumption was made on the basis of rate of web breaks occurrence. It was found that up to 64 % of total number of web breaks occurred in wet end, particularly in location of so called open draws where paper web was transferred unsupported by wire or felt. Distribution of web breaks in machine cross direction revealed that defects of paper web edge was the main reason of tearing initiation and consequent web break. The assumption was made that laser cutting was capable of improvement of laser cut edge tensile strength due to high cutting quality and sealing effect of the edge after laser cutting. Studies of laser ablation of cellulose supported this claim. Linear energy needed for cutting was calculated with regard to paper web properties in intended laser cutting location. Calculated linear cutting energy was verified with series of laser cutting. Practically obtained laser energy needed for cutting deviated from calculated values. This could be explained by difference in heat transfer via radiation in laser cutting and different absorption characteristics of dry and moist paper material. Laser cut samples (both dry and moist (dry matter content about 25-40%)) were tested for strength properties. It was shown that tensile strength and strain break of laser cut samples are similar to corresponding values of non-laser cut samples. Chosen method, however, did not address tensile strength of laser cut edge in particular. Thus, the assumption of improving strength properties with laser cutting was not fully proved. Laser cutting effect on possible pollution of mill broke (recycling of trimmed edge) was carried out. Laser cut samples (both dry and moist) were tested on the content of dirt particles. The tests revealed that accumulation of dust particles on the surface of moist samples can take place. This has to be taken into account to prevent contamination of pulp suspension when trim waste is recycled. Material loss due to evaporation during laser cutting and amount of solid residues after cutting were evaluated. Edge trimming with laser would result in 0.25 kg/h of solid residues and 2.5 kg/h of lost material due to evaporation. Schemes of laser cutting implementation and needed laser equipment were discussed. Generally, laser cutting system would require two laser sources (one laser source for each cutting zone), set of beam transfer and focusing optics and cutting heads. In order to increase reliability of system, it was suggested that each laser source would have double capacity. That would allow to perform cutting employing one laser source working at full capacity for both cutting zones. Laser technology is in required level at the moment and do not require additional development. Moreover, capacity of speed increase is high due to availability high power laser sources what can support the tendency of speed increase of paper making machines. Laser cutting system would require special roll to maintain cutting. The scheme of such roll was proposed as well as roll integration into paper making machine. Laser cutting can be done in location of central roll in press section, before so-called open draw where many web breaks occur, where it has potential to improve runability of a paper making machine. Economic performance of laser cutting was done as comparison of laser cutting system and water jet cutting working in the same conditions. It was revealed that laser cutting would still be about two times more expensive compared to water jet cutting. This is mainly due to high investment cost of laser equipment and poor energy efficiency of CO2 lasers. Another factor is that laser cutting causes material loss due to evaporation whereas water jet cutting almost does not cause material loss. Despite difficulties of laser cutting implementation in paper making machine, its implementation can be beneficial. The crucial role in that is possibility to improve cut edge strength properties and consequently reduce number of web breaks. Capacity of laser cutting to maintain cutting speeds which exceed current speeds of paper making machines what is another argument to consider laser cutting technology in design of new high speed paper making machines.

Sentence Compression with Supervised Machine Learning

The subject of the thesis is automatic sentence compression with machine learning, so that the compressed sentences remain both grammatical and retain their essential meaning. There are multiple possible uses for the compression of natural language sentences. In this thesis the focus is generation of television program subtitles, which often are compressed version of the original script of the program. The main part of the thesis consists of machine learning experiments for automatic sentence compression using different approaches to the problem. The machine learning methods used for this work are linear-chain conditional random fields and support vector machines. Also we take a look which automatic text analysis methods provide useful features for the task. The data used for machine learning is supplied by Lingsoft Inc. and consists of subtitles in both compressed an uncompressed form. The models are compared to a baseline system and comparisons are made both automatically and also using human evaluation, because of the potentially subjective nature of the output. The best result is achieved using a CRF - sequence classification using a rich feature set. All text analysis methods help classification and most useful method is morphological analysis. Tutkielman aihe on suomenkielisten lauseiden automaattinen tiivistäminen koneellisesti, niin että lyhennetyt lauseet säilyttävät olennaisen informaationsa ja pysyvät kieliopillisina. Luonnollisen kielen lauseiden tiivistämiselle on monta käyttötarkoitusta, mutta tässä tutkielmassa aihetta lähestytään television ohjelmien tekstittämisen kautta, johon käytännössä kuuluu alkuperäisen tekstin lyhentäminen televisioruudulle paremmin sopivaksi. Tutkielmassa kokeillaan erilaisia koneoppimismenetelmiä tekstin automaatiseen lyhentämiseen ja tarkastellaan miten hyvin erilaiset luonnollisen kielen analyysimenetelmät tuottavat informaatiota, joka auttaa näitä menetelmiä lyhentämään lauseita. Lisäksi tarkastellaan minkälainen lähestymistapa tuottaa parhaan lopputuloksen. Käytetyt koneoppimismenetelmät ovat tukivektorikone ja lineaarisen sekvenssin mallinen CRF. Koneoppimisen tukena käytetään tekstityksiä niiden eri käsittelyvaiheissa, jotka on saatu Lingsoft OY:ltä. Luotuja malleja vertaillaan Lopulta mallien lopputuloksia evaluoidaan automaattisesti ja koska teksti lopputuksena on jossain määrin subjektiivinen myös ihmisarviointiin perustuen. Vertailukohtana toimii kirjallisuudesta poimittu menetelmä. Tutkielman tuloksena paras lopputulos saadaan aikaan käyttäen CRF sekvenssi-luokittelijaa laajalla piirrejoukolla. Kaikki kokeillut teksin analyysimenetelmät auttavat luokittelussa, joista tärkeimmän panoksen antaa morfologinen analyysi.

Mathematical modeling of dehydration of 'Fuji' and 'Gala' apples slices using infrared

The objective of this work was to determine and model the infrared dehydration curves of apple slices - Fuji and Gala varieties. The slices were dehydrated until constant mass, in a prototype dryer with infrared heating source. The applied temperatures ranged from 50 to 100 °C. Due to the physical characteristics of the product, the dehydration curve was divided in two periods, constant and falling, separated by the critical moisture content. A linear model was used to describe the constant dehydration period. Empirical models traditionally used to model the drying behavior of agricultural products were fitted to the experimental data of the falling dehydration period. Critical moisture contents of 2.811 and 3.103 kgw kgs-1 were observed for the Fuji and Gala varieties, respectively. Based on the results, it was concluded that the constant dehydration rates presented a direct relationship with the temperature; thus, it was possible to fit a model that describes the moisture content variation in function of time and temperature. Among the tested models, which describe the falling dehydration period, the model proposed by Midilli presented the best fit for all studied conditions.

Mathematical modeling of microwave dried celery leaves and determination of the effective moisture diffusivities and activation energy

Celery (Apium graveolens L. var. secalinum Alef) leaves with 50±0.07 g weight and 91.75±0.15% humidity (~11.21 db) were dried using 8 different microwave power densities ranging between 1.8-20 W g-1, until the humidity fell down to 8.95±0.23% (~0.1 db). Microwave drying processes were completed between 5.5 and 77 min depending on the microwave power densities. In this study, measured values were compared with predicted values obtained from twenty thin layer drying theoretical, semi-empirical and empirical equations with a new thin layer drying equation. Within applied microwave power density; models whose coefficient and correlation (R²) values are highest were chosen as the best models. Weibull distribution model gave the most suitable predictions at all power density. At increasing microwave power densities, the effective moisture diffusivity values ranged from 1.595 10-10 to 6.377 10-12 m2 s-1. The activation energy was calculated using an exponential expression based on Arrhenius equation. The linear relationship between the drying rate constant and effective moisture diffusivity gave the best fit.

Modeling Energy Futures Returns With Markov Regime Switching Model

Financial time series have a tendency of abruptly changing their behavior and maintain this behavior for several consecutive periods, and commodity futures returns are not an exception. This quality proposes that nonlinear models, as opposed to linear models, can more accurately describe returns and volatility. Markov regime switching models are able to match this behavior and have become a popular way to model financial time series. This study uses Markov regime switching model to describe the behavior of energy futures returns on a commodity level, because studies show that commodity futures are a heterogeneous asset class. The purpose of this thesis is twofold. First, determine how many regimes characterize individual energy commodities’ returns in different return frequencies. Second, study the characteristics of these regimes. We extent the previous studies on the subject in two ways: We allow for the possibility that the number of regimes may exceed two, as well as conduct the research on individual commodities rather than on commodity indices or subgroups of these indices. We use daily, weekly and monthly time series of Brent crude oil, WTI crude oil, natural gas, heating oil and gasoil futures returns over 1994–2014, where available, to carry out the study. We apply the likelihood ratio test to determine the sufficient number of regimes for each commodity and data frequency. Then the time series are modeled with Markov regime switching model to obtain the return distribution characteristics of each regime, as well as the transition probabilities of moving between regimes. The results for the number of regimes suggest that daily energy futures return series consist of three to six regimes, whereas weekly and monthly returns for all energy commodities display only two regimes. When the number of regimes exceeds two, there is a tendency for the time series of energy commodities to form groups of regimes. These groups are usually quite persistent as a whole because probability of a regime switch inside the group is high. However, individual regimes in these groups are not persistent and the process oscillates between these regimes frequently. Regimes that are not part of any group are generally persistent, but show low ergodic probability, i.e. rarely prevail in the market. This study also suggests that energy futures return series characterized with two regimes do not necessarily display persistent bull and bear regimes. In fact, for the majority of time series, bearish regime is considerably less persistent. Rahoituksen aikasarjoilla on taipumus arvaamattomasti muuttaa käyttäytymistään ja jatkaa tätä uutta käyttäytymistä useiden periodien ajan, eivätkä hyödykefutuurien tuotot tee tähän poikkeusta. Tämän ominaisuuden johdosta lineaaristen mallien sijasta epälineaariset mallit pystyvät tarkemmin kuvailemaan esimerkiksi tuottojen jakauman parametreja. Markov regiiminvaihtomallit pystyvät vangitsemaan tämän ominaisuuden ja siksi niistä on tullut suosittuja rahoituksen aikasarjojen mallintamisessa. Tämä tutkimus käyttää Markov regiiminvaihtomallia kuvaamaan yksittäisten energiafutuurien tuottojen käyttäytymistä, sillä tutkimukset osoittavat hyödykefutuurien olevan hyvin heterogeeninen omaisuusluokka. Tutkimuksen tarkoitus on selvittää, kuinka monta regiimiä tarvitaan kuvaamaan energiafutuurien tuottoja eri tuottofrekvensseillä ja mitkä ovat näiden regiimien ominaisuudet. Aiempaa tutkimusta aiheesta laajennetaan määrittämällä regiimien lukumäärä tilastotieteellisen testauksen menetelmin sekä tutkimalla energiafutuureja yksittäin; ei indeksi- tai alaindeksitasolla. Tutkimuksessa käytetään päivä-, viikko- ja kuukausiaikasarjoja Brent-raakaöljyn, WTI-raakaöljyn, maakaasun, lämmitysöljyn ja polttoöljyn tuotoista aikaväliltä 1994–2014, siltä osin kuin aineistoa on saatavilla. Likelihood ratio -testin avulla estimoidaan kaikille aikasarjoille regiimien määrä,jonka jälkeen Markov regiiminvaihtomallia hyödyntäen määritetään yksittäisten regiimientuottojakaumien ominaisuudet sekä regiimien välinen transitiomatriisi. Tulokset regiimien lukumäärän osalta osoittavat, että energiafutuurien päiväkohtaisten tuottojen aikasarjoissa regiimien lukumäärä vaihtelee kolmen ja kuuden välillä. Viikko- ja kuukausituottojen kohdalla kaikkien energiafutuurien prosesseissa regiimien lukumäärä on kaksi. Kun regiimejä on enemmän kuin kaksi, on prosessilla taipumus muodostaa regiimeistä koostuvia ryhmiä. Prosessi pysyy ryhmän sisällä yleensä pitkään, koska todennäköisyys siirtyä ryhmään kuuluvien regiimien välillä on suuri. Yksittäiset regiimit ryhmän sisällä eivät kuitenkaan ole kovin pysyviä. Näin ollen prosessi vaihtelee ryhmän sisäisten regiimien välillä tiuhaan. Regiimit, jotka eivät kuulu ryhmään, ovat yleensä pysyviä, mutta prosessi ajautuu niihin vain harvoin, sillä todennäköisyys siirtyä muista regiimeistä niihin on pieni. Tutkimuksen tulokset osoittavat myös, että prosesseissa, joita ohjaa kaksi regiimiä, nämä regiimit eivät välttämättä ole pysyvät bull- ja bear-markkinatilanteet. Tulokset osoittavat sen sijaan, että bear-markkinatilanne on energiafutuureissa selvästi vähemmän pysyvä.

Dynamic modeling and simulation of the airflow cleaning device

The objective of the work is to study the flow behavior and to support the design of air cleaner by dynamic simulation.In a paper printing industry, it is necessary to monitor the quality of paper when the paper is being produced. During the production, the quality of the paper can be monitored by camera. Therefore, it is necessary to keep the camera lens clean as wood particles may fall from the paper and lie on the camera lens. In this work, the behavior of the air flow and effect of the airflow on the particles at different inlet angles are simulated. Geometries of a different inlet angles of single-channel and double-channel case were constructed using ANSYS CFD Software. All the simulations were performed in ANSYS Fluent. The simulation results of single-channel and double-channel case revealed significant differences in the behavior of the flow and the particle velocity. The main conclusion from this work are in following. 1) For the single channel case the best angle was 0 degree because in that case, the air flow can keep 60% of the particles away from the lens which would otherwise stay on lens. 2) For the double channel case, the best solution was found when the angle of the first inlet was 0 degree and the angle of second inlet was 45 degree . In that case, the airflow can keep 91% of particles away from the lens which would otherwise stay on lens.

An Eigenfunction Approach for Volatility Modeling.

In this paper, we introduce a new approach for volatility modeling in discrete and continuous time. We follow the stochastic volatility literature by assuming that the variance is a function of a state variable. However, instead of assuming that the loading function is ad hoc (e.g., exponential or affine), we assume that it is a linear combination of the eigenfunctions of the conditional expectation (resp. infinitesimal generator) operator associated to the state variable in discrete (resp. continuous) time. Special examples are the popular log-normal and square-root models where the eigenfunctions are the Hermite and Laguerre polynomials respectively. The eigenfunction approach has at least six advantages: i) it is general since any square integrable function may be written as a linear combination of the eigenfunctions; ii) the orthogonality of the eigenfunctions leads to the traditional interpretations of the linear principal components analysis; iii) the implied dynamics of the variance and squared return processes are ARMA and, hence, simple for forecasting and inference purposes; (iv) more importantly, this generates fat tails for the variance and returns processes; v) in contrast to popular models, the variance of the variance is a flexible function of the variance; vi) these models are closed under temporal aggregation.

Simulation-Based Finite-Sample Normality Tests in Linear Regressions

In the literature on tests of normality, much concern has been expressed over the problems associated with residual-based procedures. Indeed, the specialized tables of critical points which are needed to perform the tests have been derived for the location-scale model; hence reliance on available significance points in the context of regression models may cause size distortions. We propose a general solution to the problem of controlling the size normality tests for the disturbances of standard linear regression, which is based on using the technique of Monte Carlo tests.

Un outil d’évaluation neurocognitive des interactions humain-machine

De plus en plus de recherches sur les Interactions Humain-Machine (IHM) tentent d’effectuer des analyses fines de l’interaction afin de faire ressortir ce qui influence les comportements des utilisateurs. Tant au niveau de l’évaluation de la performance que de l’expérience des utilisateurs, on note qu’une attention particulière est maintenant portée aux réactions émotionnelles et cognitives lors de l’interaction. Les approches qualitatives standards sont limitées, car elles se fondent sur l’observation et des entrevues après l’interaction, limitant ainsi la précision du diagnostic. L’expérience utilisateur et les réactions émotionnelles étant de nature hautement dynamique et contextualisée, les approches d’évaluation doivent l’être de même afin de permettre un diagnostic précis de l’interaction. Cette thèse présente une approche d’évaluation quantitative et dynamique qui permet de contextualiser les réactions des utilisateurs afin d’en identifier les antécédents dans l’interaction avec un système. Pour ce faire, ce travail s’articule autour de trois axes. 1) La reconnaissance automatique des buts et de la structure de tâches de l’utilisateur, à l’aide de mesures oculométriques et d’activité dans l’environnement par apprentissage machine. 2) L’inférence de construits psychologiques (activation, valence émotionnelle et charge cognitive) via l’analyse des signaux physiologiques. 3) Le diagnostic de l‘interaction reposant sur le couplage dynamique des deux précédentes opérations. Les idées et le développement de notre approche sont illustrés par leur application dans deux contextes expérimentaux : le commerce électronique et l’apprentissage par simulation. Nous présentons aussi l’outil informatique complet qui a été implémenté afin de permettre à des professionnels en évaluation (ex. : ergonomes, concepteurs de jeux, formateurs) d’utiliser l’approche proposée pour l’évaluation d’IHM. Celui-ci est conçu de manière à faciliter la triangulation des appareils de mesure impliqués dans ce travail et à s’intégrer aux méthodes classiques d’évaluation de l’interaction (ex. : questionnaires et codage des observations).

Modeling High-Dimensional Audio Sequences with Recurrent Neural Networks

Cette thèse étudie des modèles de séquences de haute dimension basés sur des réseaux de neurones récurrents (RNN) et leur application à la musique et à la parole. Bien qu'en principe les RNN puissent représenter les dépendances à long terme et la dynamique temporelle complexe propres aux séquences d'intérêt comme la vidéo, l'audio et la langue naturelle, ceux-ci n'ont pas été utilisés à leur plein potentiel depuis leur introduction par Rumelhart et al. (1986a) en raison de la difficulté de les entraîner efficacement par descente de gradient. Récemment, l'application fructueuse de l'optimisation Hessian-free et d'autres techniques d'entraînement avancées ont entraîné la recrudescence de leur utilisation dans plusieurs systèmes de l'état de l'art. Le travail de cette thèse prend part à ce développement. L'idée centrale consiste à exploiter la flexibilité des RNN pour apprendre une description probabiliste de séquences de symboles, c'est-à-dire une information de haut niveau associée aux signaux observés, qui en retour pourra servir d'à priori pour améliorer la précision de la recherche d'information. Par exemple, en modélisant l'évolution de groupes de notes dans la musique polyphonique, d'accords dans une progression harmonique, de phonèmes dans un énoncé oral ou encore de sources individuelles dans un mélange audio, nous pouvons améliorer significativement les méthodes de transcription polyphonique, de reconnaissance d'accords, de reconnaissance de la parole et de séparation de sources audio respectivement. L'application pratique de nos modèles à ces tâches est détaillée dans les quatre derniers articles présentés dans cette thèse. Dans le premier article, nous remplaçons la couche de sortie d'un RNN par des machines de Boltzmann restreintes conditionnelles pour décrire des distributions de sortie multimodales beaucoup plus riches. Dans le deuxième article, nous évaluons et proposons des méthodes avancées pour entraîner les RNN. Dans les quatre derniers articles, nous examinons différentes façons de combiner nos modèles symboliques à des réseaux profonds et à la factorisation matricielle non-négative, notamment par des produits d'experts, des architectures entrée/sortie et des cadres génératifs généralisant les modèles de Markov cachés. Nous proposons et analysons également des méthodes d'inférence efficaces pour ces modèles, telles la recherche vorace chronologique, la recherche en faisceau à haute dimension, la recherche en faisceau élagué et la descente de gradient. Finalement, nous abordons les questions de l'étiquette biaisée, du maître imposant, du lissage temporel, de la régularisation et du pré-entraînement.

Modeling and Analysis of Some Heavy Tailed Time Series

The thesis has covered various aspects of modeling and analysis of finite mean time series with symmetric stable distributed innovations. Time series analysis based on Box and Jenkins methods are the most popular approaches where the models are linear and errors are Gaussian. We highlighted the limitations of classical time series analysis tools and explored some generalized tools and organized the approach parallel to the classical set up. In the present thesis we mainly studied the estimation and prediction of signal plus noise model. Here we assumed the signal and noise follow some models with symmetric stable innovations.We start the thesis with some motivating examples and application areas of alpha stable time series models. Classical time series analysis and corresponding theories based on finite variance models are extensively discussed in second chapter. We also surveyed the existing theories and methods correspond to infinite variance models in the same chapter. We present a linear filtering method for computing the filter weights assigned to the observation for estimating unobserved signal under general noisy environment in third chapter. Here we consider both the signal and the noise as stationary processes with infinite variance innovations. We derived semi infinite, double infinite and asymmetric signal extraction filters based on minimum dispersion criteria. Finite length filters based on Kalman-Levy filters are developed and identified the pattern of the filter weights. Simulation studies show that the proposed methods are competent enough in signal extraction for processes with infinite variance.Parameter estimation of autoregressive signals observed in a symmetric stable noise environment is discussed in fourth chapter. Here we used higher order Yule-Walker type estimation using auto-covariation function and exemplify the methods by simulation and application to Sea surface temperature data. We increased the number of Yule-Walker equations and proposed a ordinary least square estimate to the autoregressive parameters. Singularity problem of the auto-covariation matrix is addressed and derived a modified version of the Generalized Yule-Walker method using singular value decomposition.In fifth chapter of the thesis we introduced partial covariation function as a tool for stable time series analysis where covariance or partial covariance is ill defined. Asymptotic results of the partial auto-covariation is studied and its application in model identification of stable auto-regressive models are discussed. We generalize the Durbin-Levinson algorithm to include infinite variance models in terms of partial auto-covariation function and introduce a new information criteria for consistent order estimation of stable autoregressive model.In chapter six we explore the application of the techniques discussed in the previous chapter in signal processing. Frequency estimation of sinusoidal signal observed in symmetric stable noisy environment is discussed in this context. Here we introduced a parametric spectrum analysis and frequency estimate using power transfer function. Estimate of the power transfer function is obtained using the modified generalized Yule-Walker approach. Another important problem in statistical signal processing is to identify the number of sinusoidal components in an observed signal. We used a modified version of the proposed information criteria for this purpose.

ARMA modeling of time series based on rational approximation of spectral density function

This study is concerned with Autoregressive Moving Average (ARMA) models of time series. ARMA models form a subclass of the class of general linear models which represents stationary time series, a phenomenon encountered most often in practice by engineers, scientists and economists. It is always desirable to employ models which use parameters parsimoniously. Parsimony will be achieved by ARMA models because it has only finite number of parameters. Even though the discussion is primarily concerned with stationary time series, later we will take up the case of homogeneous non stationary time series which can be transformed to stationary time series. Time series models, obtained with the help of the present and past data is used for forecasting future values. Physical science as well as social science take benefits of forecasting models. The role of forecasting cuts across all fields of management-—finance, marketing, production, business economics, as also in signal process, communication engineering, chemical processes, electronics etc. This high applicability of time series is the motivation to this study.

Adaptive Real-time Anomaly-based Intrusion Detection using Data Mining and Machine Learning Techniques

Die zunehmende Vernetzung der Informations- und Kommunikationssysteme führt zu einer weiteren Erhöhung der Komplexität und damit auch zu einer weiteren Zunahme von Sicherheitslücken. Klassische Schutzmechanismen wie Firewall-Systeme und Anti-Malware-Lösungen bieten schon lange keinen Schutz mehr vor Eindringversuchen in IT-Infrastrukturen. Als ein sehr wirkungsvolles Instrument zum Schutz gegenüber Cyber-Attacken haben sich hierbei die Intrusion Detection Systeme (IDS) etabliert. Solche Systeme sammeln und analysieren Informationen von Netzwerkkomponenten und Rechnern, um ungewöhnliches Verhalten und Sicherheitsverletzungen automatisiert festzustellen. Während signatur-basierte Ansätze nur bereits bekannte Angriffsmuster detektieren können, sind anomalie-basierte IDS auch in der Lage, neue bisher unbekannte Angriffe (Zero-Day-Attacks) frühzeitig zu erkennen. Das Kernproblem von Intrusion Detection Systeme besteht jedoch in der optimalen Verarbeitung der gewaltigen Netzdaten und der Entwicklung eines in Echtzeit arbeitenden adaptiven Erkennungsmodells. Um diese Herausforderungen lösen zu können, stellt diese Dissertation ein Framework bereit, das aus zwei Hauptteilen besteht. Der erste Teil, OptiFilter genannt, verwendet ein dynamisches "Queuing Concept", um die zahlreich anfallenden Netzdaten weiter zu verarbeiten, baut fortlaufend Netzverbindungen auf, und exportiert strukturierte Input-Daten für das IDS. Den zweiten Teil stellt ein adaptiver Klassifikator dar, der ein Klassifikator-Modell basierend auf "Enhanced Growing Hierarchical Self Organizing Map" (EGHSOM), ein Modell für Netzwerk Normalzustand (NNB) und ein "Update Model" umfasst. In dem OptiFilter werden Tcpdump und SNMP traps benutzt, um die Netzwerkpakete und Hostereignisse fortlaufend zu aggregieren. Diese aggregierten Netzwerkpackete und Hostereignisse werden weiter analysiert und in Verbindungsvektoren umgewandelt. Zur Verbesserung der Erkennungsrate des adaptiven Klassifikators wird das künstliche neuronale Netz GHSOM intensiv untersucht und wesentlich weiterentwickelt. In dieser Dissertation werden unterschiedliche Ansätze vorgeschlagen und diskutiert. So wird eine classification-confidence margin threshold definiert, um die unbekannten bösartigen Verbindungen aufzudecken, die Stabilität der Wachstumstopologie durch neuartige Ansätze für die Initialisierung der Gewichtvektoren und durch die Stärkung der Winner Neuronen erhöht, und ein selbst-adaptives Verfahren eingeführt, um das Modell ständig aktualisieren zu können. Darüber hinaus besteht die Hauptaufgabe des NNB-Modells in der weiteren Untersuchung der erkannten unbekannten Verbindungen von der EGHSOM und der Überprüfung, ob sie normal sind. Jedoch, ändern sich die Netzverkehrsdaten wegen des Concept drif Phänomens ständig, was in Echtzeit zur Erzeugung nicht stationärer Netzdaten führt. Dieses Phänomen wird von dem Update-Modell besser kontrolliert. Das EGHSOM-Modell kann die neuen Anomalien effektiv erkennen und das NNB-Model passt die Änderungen in Netzdaten optimal an. Bei den experimentellen Untersuchungen hat das Framework erfolgversprechende Ergebnisse gezeigt. Im ersten Experiment wurde das Framework in Offline-Betriebsmodus evaluiert. Der OptiFilter wurde mit offline-, synthetischen- und realistischen Daten ausgewertet. Der adaptive Klassifikator wurde mit dem 10-Fold Cross Validation Verfahren evaluiert, um dessen Genauigkeit abzuschätzen. Im zweiten Experiment wurde das Framework auf einer 1 bis 10 GB Netzwerkstrecke installiert und im Online-Betriebsmodus in Echtzeit ausgewertet. Der OptiFilter hat erfolgreich die gewaltige Menge von Netzdaten in die strukturierten Verbindungsvektoren umgewandelt und der adaptive Klassifikator hat sie präzise klassifiziert. Die Vergleichsstudie zwischen dem entwickelten Framework und anderen bekannten IDS-Ansätzen zeigt, dass der vorgeschlagene IDSFramework alle anderen Ansätze übertrifft. Dies lässt sich auf folgende Kernpunkte zurückführen: Bearbeitung der gesammelten Netzdaten, Erreichung der besten Performanz (wie die Gesamtgenauigkeit), Detektieren unbekannter Verbindungen und Entwicklung des in Echtzeit arbeitenden Erkennungsmodells von Eindringversuchen.

Understanding Subsystems in Biology through Dimensionality Reduction, Graph Partitioning and Analytical Modeling

Biological systems exhibit rich and complex behavior through the orchestrated interplay of a large array of components. It is hypothesized that separable subsystems with some degree of functional autonomy exist; deciphering their independent behavior and functionality would greatly facilitate understanding the system as a whole. Discovering and analyzing such subsystems are hence pivotal problems in the quest to gain a quantitative understanding of complex biological systems. In this work, using approaches from machine learning, physics and graph theory, methods for the identification and analysis of such subsystems were developed. A novel methodology, based on a recent machine learning algorithm known as non-negative matrix factorization (NMF), was developed to discover such subsystems in a set of large-scale gene expression data. This set of subsystems was then used to predict functional relationships between genes, and this approach was shown to score significantly higher than conventional methods when benchmarking them against existing databases. Moreover, a mathematical treatment was developed to treat simple network subsystems based only on their topology (independent of particular parameter values). Application to a problem of experimental interest demonstrated the need for extentions to the conventional model to fully explain the experimental data. Finally, the notion of a subsystem was evaluated from a topological perspective. A number of different protein networks were examined to analyze their topological properties with respect to separability, seeking to find separable subsystems. These networks were shown to exhibit separability in a nonintuitive fashion, while the separable subsystems were of strong biological significance. It was demonstrated that the separability property found was not due to incomplete or biased data, but is likely to reflect biological structure.