Koulusaavutusaineiston analyysi asetelmaperusteisilla ja kaksitasoisilla logit-regressiomalleilla

Summary: The analysis of educational data with design-based and two-level logistic regression models

Coifman-aallokkeet ja skaalaussuotimen muodostus differentiaalievoluut ion avulla

Ortogonaalisen M-kaistaisen moniresoluutioanalyysin matemaattiset perusteet esitetään yksityiskohtaisesti. Coifman-aallokkeiden määritelmä yleistetään dilaatiokertoimelle M ja nollasta poikkeavalle häviävien momenttien keskukselle.Funktion approksimointia näytepisteistä aallokkeiden avulla pohditaan ja erityisesti esitetään approksimaation asymptoottinen virhearvio Coifman-aallokkeille. Skaalaussuotimelle osoitetaan välttämättömät ja riittävät ehdot, jotka johtavat yleistettyihin Coifman-aallokkeisiin. Moniresoluutioanalyysin tiheys todistetaansuoraan Lebesguen integraalin määritelmään perustuen yksikön partitio-ominaisuutta käyttäen. Todistus on riittävä sellaisenaan avaruudessa L2(Wd) käyttämättä Fourier-tason ominaisuuksia tai ehtoja. Mallatin algoritmi johdetaan M-kaistaisille aallokkeille ja moniuloitteisille signaaleille. Algoritmille esitetään myös rekursiivinen muoto. Differentiaalievoluutioalgoritmin avulla ratkaistaan Coifman-aallokkeisiin liittyvien skaalaussuotimien kertoimien arvoja useille skaalausfunktiolle. Approksimaatio- ja kuvanpakkausesimerkkejä esitetään menetelmien havainnollistamiseksi. Differentiaalievoluutioalgoritmin avulla etsitään myös referenssikuville optimoitu skaalaussuodin. Löydetty suodin on regulaarinen ja erittäinsymmetrinen.

Detection of denial of service attacks

This thesis studies techniques used for detection of distributed denial of service attacks which during last decade became one of the most serious network security threats. To evaluate different detection algorithms and further improve them we need to test their performance under conditions as close to real-life situations as possible. Currently the only feasible solution for large-scale tests is the simulated environment. The thesis describes implementation of recursive non-parametric CUSUM algorithm for detection of distributed denial of service attacks in ns-2 network simulator – a standard de-facto for network simulation.

Некоторые наблюдения по поводу построения компьютерного морфологического анализатора марийского языка. [Problems in writing a two-level model of Mari morphology]

The article describes some concrete problems that were encountered when writing a two-level model of Mari morphology. Mari is an agglutinative Finno-Ugric language spoken in Russia by about 600 000 people. The work was begun in the 1980s on the basis of K. Koskenniemi’s Two-Level Morphology (1983), but in the latest stage R. Beesley’s and L. Karttunen’s Finite State Morphology (2003) was used. Many of the problems described in the article concern the inexplicitness of the rules in Mari grammars and the lack of information about the exact distribution of some suffixes, e.g. enclitics. The Mari grammars usually give complete paradigms for a few unproblematic verb stems, whereas the difficult or unclear forms of certain verbs are only superficially discussed. Another example of phenomena that are poorly described in grammars is the way suffixes with an initial sibilant combine to stems ending in a sibilant. The help of informants and searches from electronic corpora were used to overcome such difficulties in the development of the two-level model of Mari. The variation of the order of plural markers, case suffixes and possessive suffixes is a typical feature of Mari. The morphotactic rules constructed for Mari declensional forms tend to be recursive and their productivity must be limited by some technical device, such as filters. In the present model, certain plural markers were treated like nouns. The positional and functional versatility of the possessive suffixes can be regarded as the most challenging phenomenon in attempts to formalize the Mari morphology. Cyrillic orthography, which was used in the model, also caused problems. For instance, a Cyrillic letter may represent a sequence of two sounds, the first being part of the word stem while the other belongs to a suffix. In some cases, letters for voiced consonants are also generalized to represent voiceless consonants. Such orthographical conventions distance a morphological model based on orthography from the actual (morpho)phonological processes in the language.

Demographic Modelling of Human Population Growth

This work presents models and methods that have been used in producing forecasts of population growth. The work is intended to emphasize the reliability bounds of the model forecasts. Leslie model and various versions of logistic population models are presented. References to literature and several studies are given. A lot of relevant methodology has been developed in biological sciences. The Leslie modelling approach involves the use of current trends in mortality,fertility, migration and emigration. The model treats population divided in age groups and the model is given as a recursive system. Other group of models is based on straightforward extrapolation of census data. Trajectories of simple exponential growth function and logistic models are used to produce the forecast. The work presents the basics of Leslie type modelling and the logistic models, including multi- parameter logistic functions. The latter model is also analysed from model reliability point of view. Bayesian approach and MCMC method are used to create error bounds of the model predictions.

Finanssikriisin 2007-2009 syiden identifiointi yhdysvaltalaisella aineistolla

Tutkimuksen tavoitteena on selvittää, mitkä syyt ovat vaikuttaneet Yhdysvalloissa vuonna 2007 puhjenneeseen finanssikriisiin ja kyetäänkö näistä syistä identifioimaan muuttujia, joilla pystytään ennustamaan kriisin puhkeamisen todennäköisyys. Tutkielman empiirinen osio toteutetaan logistisen regressioanalyysin avulla. Aiempaan kirjallisuuteen perustuen logistiseen regressiomallin otetaan 18 finanssikriisiä selittävää muuttujaa. Selitettävänä muuttujana on yhdysvaltalaisten pankkien tilaa kuvaava muuttuja, joka saadaan jakamalla pankkien järjestämättömät saatavat kokonaislainakannalla. Tutkimus osoitti, että aiempaan kirjallisuuteen ja aiempiin kriisejä ennustaviin malleihin perustuen kyetään logistista regressiomenetelmää käyttäen muodostamaan malli, jonka avulla löytyy tilastollisesti merkitseviä muuttujia ja joilla on vaikutusta kriisin puhkeamisen todennäköisyyteen. Kriisin puhkeamisen todennäköisyyttä lisäävien muuttujien identifiointi on haastavaa vaikka tilastollisesti merkitseviä muuttujia löytyy.

From network pictures to network insight in solution business – the role of internal communication

Operating in business-to-business markets requires an in-depth understanding on business networks. Actions and reactions made to compete in markets are fundamentally based on managers‘ subjective perceptions of the network. However, an amalgamation of these individual perceptions, termed a network picture, to a common company level shared understanding on that network, known as network insight, is found to be a substantial challenge for companies. A company‘s capability to enhance common network insight is even argued to lead competitive advantage. Especially companies with value creating logics that require wide comprehension of and collaborating in networks, such as solution business, are necessitated to develop advanced network insight. According to the extant literature, dispersed pieces of atomized network pictures can be unified to a common network insight through a process of amalgamation that comprises barriers/drivers of multilateral exchange, manifold rationality, and recursive time. However, the extant body of literature appears to lack an understanding on the role of internal communication in the development of network insight. Nonetheless, the extant understanding on the amalgamation process indicates that internal communication plays a substantial role in the development of company level network insight. The purpose of the present thesis is to enhance understanding on internal communication in the amalgamation of network pictures to develop network insight in the solution business setting, which was chosen to represent business-to-business value creating logic that emphasizes the capability to understand and utilize networks. Thus, in solution business the role of succeeding in the amalgamation process is expected to emphasize. The study combines qualitative and quantitative research by means of various analytical methods including multiple case analysis, simulation, and social network analysis. Approaching the nascent research topic with differing perspectives and means provides a broader insight on the phenomenon. The study provides empirical evidence from Finnish business-to-business companies which operate globally. The empirical data comprise interviews (n=28) with managers of three case companies. In addition the data includes a questionnaire (n=23) collected mainly for the purpose of social network analysis. In addition, the thesis includes a simulation study more specifically achieved by means of agent based modeling. The findings of the thesis shed light on the role of internal communication in the amalgamation process, contributing to the emergent discussion of network insights and thus to the industrial marketing research. In addition, the thesis increases understanding on internal communication in the change process to solution business, a supplier‘s internal communication in its matrix organization structure during a project sales process, key barriers and drivers that influence internal communication in project sales networks, perceived power within industrial project sales, and the revisioning of network pictures. According to the findings, internal communication is found to play a substantial role in the amalgamation process. First, it is suggested that internal communication is a base of multilateral exchange. Second, it is suggested that internal communication intensifies and maintains manifold rationality. Third, internal communication is needed to explicate the usually differing time perspectives of others and thus it is suggested that internal communication has role as the explicator of recursive time. Furthermore, the role of an efficient amalgamation process is found to be emphasized in solutions business as it requires a more advanced network insight for cross-functional collaboration. Finally, the thesis offers several managerial implications for industrial suppliers to enhance the amalgamation process when operating in solution business.

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.

Kalman filter for computational market dynamics

Kalman filter is a recursive mathematical power tool that plays an increasingly vital role in innumerable fields of study. The filter has been put to service in a multitude of studies involving both time series modelling and financial time series modelling. Modelling time series data in Computational Market Dynamics (CMD) can be accomplished using the Jablonska-Capasso-Morale (JCM) model. Maximum likelihood approach has always been utilised to estimate the parameters of the JCM model. The purpose of this study is to discover if the Kalman filter can be effectively utilized in CMD. Ensemble Kalman filter (EnKF), with 50 ensemble members, applied to US sugar prices spanning the period of January, 1960 to February, 2012 was employed for this work. The real data and Kalman filter trajectories showed no significant discrepancies, hence indicating satisfactory performance of the technique. Since only US sugar prices were utilized, it would be interesting to discover the nature of results if other data sets are employed.

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

Real-time simulation of multibody systems with applications for working mobile vehicles

This dissertation describes an approach for developing a real-time simulation for working mobile vehicles based on multibody modeling. The use of multibody modeling allows comprehensive description of the constrained motion of the mechanical systems involved and permits real-time solving of the equations of motion. By carefully selecting the multibody formulation method to be used, it is possible to increase the accuracy of the multibody model while at the same time solving equations of motion in real-time. In this study, a multibody procedure based on semi-recursive and augmented Lagrangian methods for real-time dynamic simulation application is studied in detail. In the semirecursive approach, a velocity transformation matrix is introduced to describe the dependent coordinates into relative (joint) coordinates, which reduces the size of the generalized coordinates. The augmented Lagrangian method is based on usage of global coordinates and, in that method, constraints are accounted using an iterative process. A multibody system can be modelled as either rigid or flexible bodies. When using flexible bodies, the system can be described using a floating frame of reference formulation. In this method, the deformation mode needed can be obtained from the finite element model. As the finite element model typically involves large number of degrees of freedom, reduced number of deformation modes can be obtained by employing model order reduction method such as Guyan reduction, Craig-Bampton method and Krylov subspace as shown in this study The constrained motion of the working mobile vehicles is actuated by the force from the hydraulic actuator. In this study, the hydraulic system is modeled using lumped fluid theory, in which the hydraulic circuit is divided into volumes. In this approach, the pressure wave propagation in the hoses and pipes is neglected. The contact modeling is divided into two stages: contact detection and contact response. Contact detection determines when and where the contact occurs, and contact response provides the force acting at the collision point. The friction between tire and ground is modelled using the LuGre friction model, which describes the frictional force between two surfaces. Typically, the equations of motion are solved in the full matrices format, where the sparsity of the matrices is not considered. Increasing the number of bodies and constraint equations leads to the system matrices becoming large and sparse in structure. To increase the computational efficiency, a technique for solution of sparse matrices is proposed in this dissertation and its implementation demonstrated. To assess the computing efficiency, augmented Lagrangian and semi-recursive methods are implemented employing a sparse matrix technique. From the numerical example, the results show that the proposed approach is applicable and produced appropriate results within the real-time period.

Cohesion Metrics for Improving Software Quality

Abstract Software product metrics aim at measuring the quality of software. Modu- larity is an essential factor in software quality. In this work, metrics related to modularity and especially cohesion of the modules, are considered. The existing metrics are evaluated, and several new alternatives are proposed. The idea of cohesion of modules is that a module or a class should consist of related parts. The closely related principle of coupling says that the relationships between modules should be minimized. First, internal cohesion metrics are considered. The relations that are internal to classes are shown to be useless for quality measurement. Second, we consider external relationships for cohesion. A detailed analysis using design patterns and refactorings confirms that external cohesion is a better quality indicator than internal. Third, motivated by the successes (and problems) of external cohesion metrics, another kind of metric is proposed that represents the quality of modularity of software. This metric can be applied to refactorings related to classes, resulting in a refactoring suggestion system. To describe the metrics formally, a notation for programs is developed. Because of the recursive nature of programming languages, the properties of programs are most compactly represented using grammars and formal lan- guages. Also the tools that were used for metrics calculation are described.

Optimization and Applications of Large-Scale Biomedical Event Networks

The overwhelming amount and unprecedented speed of publication in the biomedical domain make it difficult for life science researchers to acquire and maintain a broad view of the field and gather all information that would be relevant for their research. As a response to this problem, the BioNLP (Biomedical Natural Language Processing) community of researches has emerged and strives to assist life science researchers by developing modern natural language processing (NLP), information extraction (IE) and information retrieval (IR) methods that can be applied at large-scale, to scan the whole publicly available biomedical literature and extract and aggregate the information found within, while automatically normalizing the variability of natural language statements. Among different tasks, biomedical event extraction has received much attention within BioNLP community recently. Biomedical event extraction constitutes the identification of biological processes and interactions described in biomedical literature, and their representation as a set of recursive event structures. The 2009–2013 series of BioNLP Shared Tasks on Event Extraction have given raise to a number of event extraction systems, several of which have been applied at a large scale (the full set of PubMed abstracts and PubMed Central Open Access full text articles), leading to creation of massive biomedical event databases, each of which containing millions of events. Sinece top-ranking event extraction systems are based on machine-learning approach and are trained on the narrow-domain, carefully selected Shared Task training data, their performance drops when being faced with the topically highly varied PubMed and PubMed Central documents. Specifically, false-positive predictions by these systems lead to generation of incorrect biomolecular events which are spotted by the end-users. This thesis proposes a novel post-processing approach, utilizing a combination of supervised and unsupervised learning techniques, that can automatically identify and filter out a considerable proportion of incorrect events from large-scale event databases, thus increasing the general credibility of those databases. The second part of this thesis is dedicated to a system we developed for hypothesis generation from large-scale event databases, which is able to discover novel biomolecular interactions among genes/gene-products. We cast the hypothesis generation problem as a supervised network topology prediction, i.e predicting new edges in the network, as well as types and directions for these edges, utilizing a set of features that can be extracted from large biomedical event networks. Routine machine learning evaluation results, as well as manual evaluation results suggest that the problem is indeed learnable. This work won the Best Paper Award in The 5th International Symposium on Languages in Biology and Medicine (LBM 2013).