Fuzzy Principal Components Analysis

Due to the large number of characteristics, there is a need to extract the most relevant characteristicsfrom the input data, so that the amount of information lost in this way is minimal, and the classification realized with the projected data set is relevant with respect to the original data. In order to achieve this feature extraction, different statistical techniques, as well as the principal components analysis (PCA) may be used. This thesis describes an extension of principal components analysis (PCA) allowing the extraction ofa finite number of relevant features from high-dimensional fuzzy data and noisy data. PCA finds linear combinations of the original measurement variables that describe the significant variation in the data. The comparisonof the two proposed methods was produced by using postoperative patient data. Experiment results demonstrate the ability of using the proposed two methods in complex data. Fuzzy PCA was used in the classificationproblem. The classification was applied by using the similarity classifier algorithm where total similarity measures weights are optimized with differential evolution algorithm. This thesis presents the comparison of the classification results based on the obtained data from the fuzzy PCA.

Predicting prognosis using molecular profiling in estrogen receptor-positive breast cancer treated with tamoxifen.

Background: Estrogen receptor positive (ER+) breast cancers (BC) are heterogeneous with regard to their clinical behavior and response to therapies. The ER is currently the best predictor of response to the anti-estrogen agent tamoxifen, yet up to 30-40% of ER+ BC will relapse despite tamoxifen treatment. New prognostic biomarkers and further biological understanding of tamoxifen resistance are required. We used gene expression profiling to develop an outcome-based predictor using a training set of 255 ER+ BC samples from women treated with adjuvant tamoxifen monotherapy. We used clusters of highly correlated genes to develop our predictor to facilitate both signature stability and biological interpretation. Independent validation was performed using 362 tamoxifen-treated ER+ BC samples obtained from multiple institutions and treated with tamoxifen only in the adjuvant and metastatic settings.Results: We developed a gene classifier consisting of 181 genes belonging to 13 biological clusters. In the independent set of adjuvantly-treated samples, it was able to define two distinct prognostic groups (HR 2.01 95% CI: 1.29-3.13; p = 0.002). Six of the 13 gene clusters represented pathways involved in cell cycle and proliferation. In 112 metastatic breast cancer patients treated with tamoxifen, one of the classifier components suggesting a cellular inflammatory mechanism was significantly predictive of response.Conclusion: We have developed a gene classifier that can predict clinical outcome in tamoxifen-treated ER+ BC patients. Whilst our study emphasizes the important role of proliferation genes in prognosis, our approach proposes other genes and pathways that may elucidate further mechanisms that influence clinical outcome and prediction of response to tamoxifen.

Visual measurement of laser-arc hybrid welding

Tässä työssä raportoidaan hybridihitsauksesta otettujen suurnopeuskuvasarjojen automaattisen analyysijärjestelmän kehittäminen.Järjestelmän tarkoitus oli tuottaa tietoa, joka avustaisi analysoijaa arvioimaan kuvatun hitsausprosessin laatua. Tutkimus keskittyi valokaaren taajuuden säännöllisyyden ja lisäainepisaroiden lentosuuntien mittaamiseen. Valokaaria havaittiin kuvasarjoista sumean c-means-klusterointimenetelmän avullaja perättäisten valokaarien välistä aikaväliä käytettiin valokaaren taajuuden säännöllisyyden mittarina. Pisaroita paikannettiin menetelmällä, jossa yhdistyi pääkomponenttianalyysi ja tukivektoriluokitin. Kalman-suodinta käytettiin tuottamaan arvioita pisaroiden lentosuunnista ja nopeuksista. Lentosuunnanmääritysmenetelmä luokitteli pisarat niiden arvioitujen lentosuuntien perusteella. Järjestelmän kehittämiseen käytettävissä olleet kuvasarjat poikkesivat merkittävästi toisistaan kuvanlaadun ja pisaroiden ulkomuodon osalta, johtuen eroista kuvaus- ja hitsausprosesseissa. Analyysijärjestelmä kehitettiin toimimaan pienellä osajoukolla kuvasarjoja, joissa oli tietynlainen kuvaus- ja hitsausprosessi ja joiden kuvanlaatu ja pisaroiden ulkomuoto olivat samankaltaisia, mutta järjestelmää testattiin myös osajoukon ulkopuolisilla kuvasarjoilla. Testitulokset osoittivat, että lentosuunnanmääritystarkkuus oli kohtuullisen suuri osajoukonsisällä ja pieni muissa kuvasarjoissa. Valokaaren taajuuden säännöllisyyden määritys oli tarkka useammassa kuvasarjassa.

Identification of a Poor-Prognosis BRAF-Mutant-Like Population of Patients With Colon Cancer.

PURPOSE Our purpose was development and assessment of a BRAF-mutant gene expression signature for colon cancer (CC) and the study of its prognostic implications. Materials and METHODS A set of 668 stage II and III CC samples from the PETACC-3 (Pan-European Trails in Alimentary Tract Cancers) clinical trial were used to assess differential gene expression between c.1799T>A (p.V600E) BRAF mutant and non-BRAF, non-KRAS mutant cancers (double wild type) and to construct a gene expression-based classifier for detecting BRAF mutant samples with high sensitivity. The classifier was validated in independent data sets, and survival rates were compared between classifier positive and negative tumors. Results A 64 gene-based classifier was developed with 96% sensitivity and 86% specificity for detecting BRAF mutant tumors in PETACC-3 and independent samples. A subpopulation of BRAF wild-type patients (30% of KRAS mutants, 13% of double wild type) showed a gene expression pattern and had poor overall survival and survival after relapse, similar to those observed in BRAF-mutant patients. Thus they form a distinct prognostic subgroup within their mutation class. CONCLUSION A characteristic pattern of gene expression is associated with and accurately predicts BRAF mutation status and, in addition, identifies a population of BRAF mutated-like KRAS mutants and double wild-type patients with similarly poor prognosis. This suggests a common biology between these tumors and provides a novel classification tool for cancers, adding prognostic and biologic information that is not captured by the mutation status alone. These results may guide therapeutic strategies for this patient segment and may help in population stratification for clinical trials.

Modules d'endo-permutation

Dans la th´eorie des repr´esentations modulaires des groupes finis, les modules d?endo-permutation occupent une place importante. En e_et, c?est le r?ole jou´e par ces modules dans l?analyse de la structure de certains modules simples pour des groupes finis p-nilpotents, qui a amen´e E. Dade `a en introduire le concept, en 1978. Quelques ann´ees plus tard, L. Puig a d´emontr´e que la source de n?importe quel module simple pour un groupe fini p-r´esoluble quelconque est un module d?endo-permutation. Plus r´ecemment, on s?est rendu compte que ces modules interviennent aussi dans l?analyse locale des cat´egories d´eriv´ees et dans l?´etude des syst`emes de fusion. La situation que l?on consid`ere est la suivante. On se donne un nombre premier p, un p-groupe fini P, un corps alg´ebriquement clos k de caract´eristique p et on veut d´eterminer tous les kP-modules d?endo-permutation couverts ind´ecomposables de type fini, c?est-`a-dire tous les kP-modules ind´ecomposables de type fini, tels que leur alg`ebre d?endomorphismes est un kP-module de permutation ayant un facteur direct trivial. On d´efinit une relation d?´equivalence sur l?ensemble de ces kP-modules et le produit tensoriel des modules induit une structure de groupe ab´elien sur l?ensemble des classes d?´equivalence. On appelle ce groupe, le groupe de Dade de P. Ainsi, classifier les modules d?endo-permutation couverts revient `a d´eterminer le groupe de Dade de P. Le groupe de Dade d?un p-groupe fini arbitraire est encore inconnu, bien qu?E. Dade, en 1978, ´etait d´ej`a parvenu `a la classification dans le cas o`u P est ab´elien. La premi`ere partie de ce travail de th`ese est consacr´ee au probl`eme de la classification dans le cas g´en´eral et r´esoud la question dans le cas de deux familles de p-groupes finis, `a savoir celle des p-groupes m´etacycliques, pour un nombre premier p impair, et celle des 2-groupes extrasp´eciaux, de la forme D8 _ · · · _ D8. Ces deux choix ont ´et´e motiv´es par le fait que ces groupes sont "presque" ab´eliens. De plus, certains r´esultats sur la structure du groupe de Dade d?un p-groupe fini quelconque rendent le groupe de Dade des groupes de ces deux familles plus simple `a ´etudier. Dans un deuxi`eme temps, nous nous sommes int´eress´es `a deux occurrences de ces modules dans la th´eorie de la repr´esentation des groupes finis, c?est-`a-dire `a deux raisons qui motivent leur ´etude. Ainsi, nous avons r´ealis´e des modules d?endo-permutation comme sources de modules simples. En particulier, il s?av`ere que, dans le cas d?un nombre premier p impair, tout module d?endo-permutation ind´ecomposable dont la classe est un ´el´ement de torsion dans le groupe de Dade est la source d?un module simple. Finalement, nous avons d´etermin´e, parmi tous les modules d?endo-permutation connus actuellement, lesquels poss`edent une r´esolution de permutation endo-scind´ee. Nous sommes arriv´es `a la conclusion que les seuls modules d?endo-permutation qui n?ont pas de r´esolution de permutation endo-scind´ee sont les modules "exceptionnels" apparaissant pour un 2-groupe de quaternions g´en´eralis´es.<br/><br/>In modular representation theory, endo-permutation modules occupy an important position. Indeed, the role that these modules play, in the analysis of the structure of some particular simple modules for finite p-nilpotent groups, induced E. Dade, in 1978, to give them their current name. A few years later, L. Puig proved that the source of any simple module for any finite psolvable group is an endo-permutation module. More recently, the occurrence of endo-permutation modules has also been noticed in the local analysis of splendid equivalences between derived categories and in the study of fusion systems. We consider the following situation. Given a prime number p, a finite pgroup P and an algebraically closed field k of characteristic p, we are looking for all finitely generated indecomposable capped endo-permutation kP-modules. That is, all finitely generated indecomposable kP-modules such that their endomorphism algebra is a permutation kP-module having a trivial direct summand. Then, we define an equivalence relation on the set of all isomorphism classes of such modules, and it turns out that the tensor product (over k) induces a structure of abelian group on this set. We call this group the Dade group of P. Hence, classifying all indecomposable finitely generated capped endo-permutation kPmodules is equivalent to determining the Dade group of P. At present, the Dade group of an arbitrary finite p-group is still unknown. However, E. Dade computed the Dade group of all finite abelian p-groups, in 1978 already. The first part of this doctoral thesis is concerned with the problem of the classification in the general case and solve it in the case of two families of finite p-groups, namely the metacyclic p-groups, for an odd prime number p, and the extraspecial 2-groups of the shape D8 _· · ·_D8. These two choices have been motivated by the fact that these groups are not far from being abelian. Moreover, some general results concerning the Dade group of arbitrary finite p-groups suggest that the Dade group of the groups belonging to these two families is easier to study. In the second part of this thesis, we have been looking at two particular occurrences of these modules in representation theory of finite groups which motivate the interest of their classification. Thus, we realised endo-permutation modules as sources of simple modules. In particular, it turns out that, in case p is an odd prime, any indecomposable module whose class in the Dade group is a torsion element is the source of some simple module. Finally, we considered all the modules we know at present and determined which ones have an endo-split permutation resolution. We could then conclude that all but the "exceptionnal" modules occurring in the generalized quaternion case have an endo-split permutation resolution.<br/><br/>"Module d?endo-permutation" n?est pas le nom d?une maladie exotique contagieuse (du moins pas `a ma connaissance), comme vous pourriez peut-?etre l?imaginer si vous faites partie des personnes qui croient que le titre de docteur n?est destin´e qu?aux m´edecins. Dans ce cas, il se peut que le sujet dont il est question ici vous cause quelques naus´ees et r´eveille de douloureux souvenirs d?´ecole, car un module d?endo-permutation est un objet math´ematique, alg´ebrique, plus pr´ecis´ement. Ce concept a ´et´e introduit il y a un quart de si`ecle, de l?autre c?ot´e de l?Atlantique, et il s?est r´ev´el´e su_samment int´eressant pour qu?aujourd?hui il ait franchi bien des fronti`eres, celles de l?alg`ebre y compris. Mais de quoi s?agit-il ? Si vous entendez le terme "endo-permutation" probablement pour la premi`ere fois, ce n?est certainement pas le cas pour celui de "module". Cependant, sa d´efinition dans le pr´esent contexte ne co¨ýncide avec aucune de celles figurant dans les dictionnaires ordinaires. Les personnes qui ont d´ej`a entendu parler de Frobenius, Burnside, Schur, ou encore Brauer, pourront vous dire qu?un module est une repr´esentation. "De quoi ?" vous demanderezvous. "Un spectacle de marionnettes, peut-?etre ?" Bien s?ur que non ! Un module d?endo-permutation est une repr´esentation particuli`ere de certains groupes finis, o`u un groupe n?est pas un groupe de rock, comme vous pouvez vous en douter, mais d´esigne un objet math´ematique connu par tous les ´etudiants en sciences au terme de leur premi`ere ann´ee universitaire (en th´eorie, du moins). La "popularit´e" de la notion de groupe, fini ou non, est due au fait que les groupes sont fr´equemment utilis´es, aussi bien dans le domaine abstrait des math´ematiques, que dans le monde r´eel des physiciens, chimistes et autres biologistes (pour ne citer qu?eux). "Mais comment peut-on utiliser concr`etement ces objets invisibles ?" vous demanderez-vous alors. Et bien, justement, en les consid´erant par l?interm´ediaire de leurs repr´esentations, c?est-`a-dire en leur associant des matrices, de fa¸con plus ou moins naturelle. Or, comme il y a "beaucoup trop" de matrices pour un groupe donn´e, elles sont classifi´ees selon certaines de leurs propri´et´es, ce qui permet de les r´epertorier dans diverses familles (celle des modules d?endo-permutation, par exemple). Un groupe est ainsi rendu "concret", car les donn´ees matricielles sont manipulables par tous les scienti- fiques (et leurs ordinateurs), qui peuvent alors les utiliser dans leurs recherches, afin de contribuer au progr`es de la science. En toute franchise, c?est bien loin de ces soucis terre-`a-terre que ce travail de th`ese sur la classification des modules d?endo-permutation a ´et´e accompli. En fait, quitte `a choquer certaines ?ames sensibles, sa r´ealisation est surtout due au caract`ere ´epicure de son auteur, qui, avouons-le, en a ´et´e pleinement satisfait !

Integral cohomology of finite postnikov towers

Depuis le séminaire H. Cartan de 1954-55, il est bien connu que l'on peut trouver des éléments de torsion arbitrairement grande dans l'homologie entière des espaces d'Eilenberg-MacLane K(G,n) où G est un groupe abélien non trivial et n>1. L'objectif majeur de ce travail est d'étendre ce résultat à des H-espaces possédant plus d'un groupe d'homotopie non trivial. Dans le but de contrôler précisément le résultat de H. Cartan, on commence par étudier la dualité entre l'homologie et la cohomologie des espaces d'Eilenberg-MacLane 2-locaux de type fini. On parvient ainsi à raffiner quelques résultats qui découlent des calculs de H. Cartan. Le résultat principal de ce travail peut être formulé comme suit. Soit X un H-espace ne possédant que deux groupes d'homotopie non triviaux, tous deux finis et de 2-torsion. Alors X n'admet pas d'exposant pour son groupe gradué d'homologie entière réduite. On construit une large classe d'espaces pour laquelle ce résultat n'est qu'une conséquence d'une caractéristique topologique, à savoir l'existence d'un rétract faible X K(G,n) pour un certain groupe abélien G et n>1. On généralise également notre résultat principal à des espaces plus compliqués en utilisant la suite spectrale d'Eilenberg-Moore ainsi que des méthodes analytiques faisant apparaître les nombres de Betti et leur comportement asymptotique. Finalement, on conjecture que les espaces qui ne possédent qu'un nombre fini de groupes d'homotopie non triviaux n'admettent pas d'exposant homologique. Ce travail contient par ailleurs la présentation de la « machine d'Eilenberg-MacLane », un programme C++ conçu pour calculer explicitement les groupes d'homologie entière des espaces d'Eilenberg-MacLane. <br/><br/>By the work of H. Cartan, it is well known that one can find elements of arbitrarilly high torsion in the integral (co)homology groups of an Eilenberg-MacLane space K(G,n), where G is a non-trivial abelian group and n>1. The main goal of this work is to extend this result to H-spaces having more than one non-trivial homotopy groups. In order to have an accurate hold on H. Cartan's result, we start by studying the duality between homology and cohomology of 2-local Eilenberg-MacLane spaces of finite type. This leads us to some improvements of H. Cartan's methods in this particular case. Our main result can be stated as follows. Let X be an H-space with two non-vanishing finite 2-torsion homotopy groups. Then X does not admit any exponent for its reduced integral graded (co)homology group. We construct a wide class of examples for which this result is a simple consequence of a topological feature, namely the existence of a weak retract X K(G,n) for some abelian group G and n>1. We also generalize our main result to more complicated stable two stage Postnikov systems, using the Eilenberg-Moore spectral sequence and analytic methods involving Betti numbers and their asymptotic behaviour. Finally, we investigate some guesses on the non-existence of homology exponents for finite Postnikov towers. We conjecture that Postnikov pieces do not admit any (co)homology exponent. This work also includes the presentation of the "Eilenberg-MacLane machine", a C++ program designed to compute explicitely all integral homology groups of Eilenberg-MacLane spaces. <br/><br/>Il est toujours difficile pour un mathématicien de parler de son travail. La difficulté réside dans le fait que les objets qu'il étudie sont abstraits. On rencontre assez rarement un espace vectoriel, une catégorie abélienne ou une transformée de Laplace au coin de la rue ! Cependant, même si les objets mathématiques sont difficiles à cerner pour un non-mathématicien, les méthodes pour les étudier sont essentiellement les mêmes que celles utilisées dans les autres disciplines scientifiques. On décortique les objets complexes en composantes plus simples à étudier. On dresse la liste des propriétés des objets mathématiques, puis on les classe en formant des familles d'objets partageant un caractère commun. On cherche des façons différentes, mais équivalentes, de formuler un problème. Etc. Mon travail concerne le domaine mathématique de la topologie algébrique. Le but ultime de cette discipline est de parvenir à classifier tous les espaces topologiques en faisant usage de l'algèbre. Cette activité est comparable à celle d'un ornithologue (topologue) qui étudierait les oiseaux (les espaces topologiques) par exemple à l'aide de jumelles (l'algèbre). S'il voit un oiseau de petite taille, arboricole, chanteur et bâtisseur de nids, pourvu de pattes à quatre doigts, dont trois en avant et un, muni d'une forte griffe, en arrière, alors il en déduira à coup sûr que c'est un passereau. Il lui restera encore à déterminer si c'est un moineau, un merle ou un rossignol. Considérons ci-dessous quelques exemples d'espaces topologiques: a) un cube creux, b) une sphère et c) un tore creux (c.-à-d. une chambre à air). a) b) c) Si toute personne normalement constituée perçoit ici trois figures différentes, le topologue, lui, n'en voit que deux ! De son point de vue, le cube et la sphère ne sont pas différents puisque ils sont homéomorphes: on peut transformer l'un en l'autre de façon continue (il suffirait de souffler dans le cube pour obtenir la sphère). Par contre, la sphère et le tore ne sont pas homéomorphes: triturez la sphère de toutes les façons (sans la déchirer), jamais vous n'obtiendrez le tore. Il existe un infinité d'espaces topologiques et, contrairement à ce que l'on serait naïvement tenté de croire, déterminer si deux d'entre eux sont homéomorphes est très difficile en général. Pour essayer de résoudre ce problème, les topologues ont eu l'idée de faire intervenir l'algèbre dans leurs raisonnements. Ce fut la naissance de la théorie de l'homotopie. Il s'agit, suivant une recette bien particulière, d'associer à tout espace topologique une infinité de ce que les algébristes appellent des groupes. Les groupes ainsi obtenus sont appelés groupes d'homotopie de l'espace topologique. Les mathématiciens ont commencé par montrer que deux espaces topologiques qui sont homéomorphes (par exemple le cube et la sphère) ont les même groupes d'homotopie. On parle alors d'invariants (les groupes d'homotopie sont bien invariants relativement à des espaces topologiques qui sont homéomorphes). Par conséquent, deux espaces topologiques qui n'ont pas les mêmes groupes d'homotopie ne peuvent en aucun cas être homéomorphes. C'est là un excellent moyen de classer les espaces topologiques (pensez à l'ornithologue qui observe les pattes des oiseaux pour déterminer s'il a affaire à un passereau ou non). Mon travail porte sur les espaces topologiques qui n'ont qu'un nombre fini de groupes d'homotopie non nuls. De tels espaces sont appelés des tours de Postnikov finies. On y étudie leurs groupes de cohomologie entière, une autre famille d'invariants, à l'instar des groupes d'homotopie. On mesure d'une certaine manière la taille d'un groupe de cohomologie à l'aide de la notion d'exposant; ainsi, un groupe de cohomologie possédant un exposant est relativement petit. L'un des résultats principaux de ce travail porte sur une étude de la taille des groupes de cohomologie des tours de Postnikov finies. Il s'agit du théorème suivant: un H-espace topologique 1-connexe 2-local et de type fini qui ne possède qu'un ou deux groupes d'homotopie non nuls n'a pas d'exposant pour son groupe gradué de cohomologie entière réduite. S'il fallait interpréter qualitativement ce résultat, on pourrait dire que plus un espace est petit du point de vue de la cohomologie (c.-à-d. s'il possède un exposant cohomologique), plus il est intéressant du point de vue de l'homotopie (c.-à-d. il aura plus de deux groupes d'homotopie non nuls). Il ressort de mon travail que de tels espaces sont très intéressants dans le sens où ils peuvent avoir une infinité de groupes d'homotopie non nuls. Jean-Pierre Serre, médaillé Fields en 1954, a montré que toutes les sphères de dimension >1 ont une infinité de groupes d'homotopie non nuls. Des espaces avec un exposant cohomologique aux sphères, il n'y a qu'un pas à franchir...

Studies on integrating kinematic design method with mechanical systems simulation techniques

Theultimate goal of any research in the mechanism/kinematic/design area may be called predictive design, ie the optimisation of mechanism proportions in the design stage without requiring extensive life and wear testing. This is an ambitious goal and can be realised through development and refinement of numerical (computational) technology in order to facilitate the design analysis and optimisation of complex mechanisms, mechanical components and systems. As a part of the systematic design methodology this thesis concentrates on kinematic synthesis (kinematic design and analysis) methods in the mechanism synthesis process. The main task of kinematic design is to find all possible solutions in the form of structural parameters to accomplish the desired requirements of motion. Main formulations of kinematic design can be broadly divided to exact synthesis and approximate synthesis formulations. The exact synthesis formulation is based in solving n linear or nonlinear equations in n variables and the solutions for the problem areget by adopting closed form classical or modern algebraic solution methods or using numerical solution methods based on the polynomial continuation or homotopy. The approximate synthesis formulations is based on minimising the approximation error by direct optimisation The main drawbacks of exact synthesis formulationare: (ia) limitations of number of design specifications and (iia) failure in handling design constraints- especially inequality constraints. The main drawbacks of approximate synthesis formulations are: (ib) it is difficult to choose a proper initial linkage and (iib) it is hard to find more than one solution. Recentformulations in solving the approximate synthesis problem adopts polynomial continuation providing several solutions, but it can not handle inequality const-raints. Based on the practical design needs the mixed exact-approximate position synthesis with two exact and an unlimited number of approximate positions has also been developed. The solutions space is presented as a ground pivot map but thepole between the exact positions cannot be selected as a ground pivot. In this thesis the exact synthesis problem of planar mechanism is solved by generating all possible solutions for the optimisation process ¿ including solutions in positive dimensional solution sets - within inequality constraints of structural parameters. Through the literature research it is first shown that the algebraic and numerical solution methods ¿ used in the research area of computational kinematics ¿ are capable of solving non-parametric algebraic systems of n equations inn variables and cannot handle the singularities associated with positive-dimensional solution sets. In this thesis the problem of positive-dimensional solutionsets is solved adopting the main principles from mathematical research area of algebraic geometry in solving parametric ( in the mathematical sense that all parameter values are considered ¿ including the degenerate cases ¿ for which the system is solvable ) algebraic systems of n equations and at least n+1 variables.Adopting the developed solution method in solving the dyadic equations in direct polynomial form in two- to three-precision-points it has been algebraically proved and numerically demonstrated that the map of the ground pivots is ambiguousand that the singularities associated with positive-dimensional solution sets can be solved. The positive-dimensional solution sets associated with the poles might contain physically meaningful solutions in the form of optimal defectfree mechanisms. Traditionally the mechanism optimisation of hydraulically driven boommechanisms is done at early state of the design process. This will result in optimal component design rather than optimal system level design. Modern mechanismoptimisation at system level demands integration of kinematic design methods with mechanical system simulation techniques. In this thesis a new kinematic design method for hydraulically driven boom mechanism is developed and integrated in mechanical system simulation techniques. The developed kinematic design method is based on the combinations of two-precision-point formulation and on optimisation ( with mathematical programming techniques or adopting optimisation methods based on probability and statistics ) of substructures using calculated criteria from the system level response of multidegree-of-freedom mechanisms. Eg. by adopting the mixed exact-approximate position synthesis in direct optimisation (using mathematical programming techniques) with two exact positions and an unlimitednumber of approximate positions the drawbacks of (ia)-(iib) has been cancelled.The design principles of the developed method are based on the design-tree -approach of the mechanical systems and the design method ¿ in principle ¿ is capable of capturing the interrelationship between kinematic and dynamic synthesis simultaneously when the developed kinematic design method is integrated with the mechanical system simulation techniques.

Dynamic Model of a Small Once-Through Boiler

In this study, a model for the unsteady dynamic behaviour of a once-through counter flow boiler that uses an organic working fluid is presented. The boiler is a compact waste-heat boiler without a furnace and it has a preheater, a vaporiser and a superheater. The relative lengths of the boiler parts vary with the operating conditions since they are all parts of a single tube. The present research is a part of a study on the unsteady dynamics of an organic Rankine cycle power plant and it will be a part of a dynamic process model. The boiler model is presented using a selected example case that uses toluene as the process fluid and flue gas from natural gas combustion as the heat source. The dynamic behaviour of the boiler means transition from the steady initial state towards another steady state that corresponds to the changed process conditions. The solution method chosen was to find such a pressure of the process fluid that the mass of the process fluid in the boiler equals the mass calculated using the mass flows into and out of the boiler during a time step, using the finite difference method. A special method of fast calculation of the thermal properties has been used, because most of the calculation time is spent in calculating the fluid properties. The boiler was divided into elements. The values of the thermodynamic properties and mass flows were calculated in the nodes that connect the elements. Dynamic behaviour was limited to the process fluid and tube wall, and the heat source was regarded as to be steady. The elements that connect the preheater to thevaporiser and the vaporiser to the superheater were treated in a special way that takes into account a flexible change from one part to the other. The model consists of the calculation of the steady state initial distribution of the variables in the nodes, and the calculation of these nodal values in a dynamic state. The initial state of the boiler was received from a steady process model that isnot a part of the boiler model. The known boundary values that may vary during the dynamic calculation were the inlet temperature and mass flow rates of both the heat source and the process fluid. A brief examination of the oscillation around a steady state, the so-called Ledinegg instability, was done. This examination showed that the pressure drop in the boiler is a third degree polynomial of the mass flow rate, and the stability criterion is a second degree polynomial of the enthalpy change in the preheater. The numerical examination showed that oscillations did not exist in the example case. The dynamic boiler model was analysed for linear and step changes of the entering fluid temperatures and flow rates.The problem for verifying the correctness of the achieved results was that there was no possibility o compare them with measurements. This is why the only way was to determine whether the obtained results were intuitively reasonable and the results changed logically when the boundary conditions were changed. The numerical stability was checked in a test run in which there was no change in input values. The differences compared with the initial values were so small that the effects of numerical oscillations were negligible. The heat source side tests showed that the model gives results that are logical in the directions of the changes, and the order of magnitude of the timescale of changes is also as expected. The results of the tests on the process fluid side showed that the model gives reasonable results both on the temperature changes that cause small alterations in the process state and on mass flow rate changes causing very great alterations. The test runs showed that the dynamic model has no problems in calculating cases in which temperature of the entering heat source suddenly goes below that of the tube wall or the process fluid.

Supervised localimage feature detection

This thesis is about detection of local image features. The research topic belongs to the wider area of object detection, which is a machine vision and pattern recognition problem where an object must be detected (located) in an image. State-of-the-art object detection methods often divide the problem into separate interest point detection and local image description steps, but in this thesis a different technique is used, leading to higher quality image features which enable more precise localization. Instead of using interest point detection the landmark positions are marked manually. Therefore, the quality of the image features is not limited by the interest point detection phase and the learning of image features is simplified. The approach combines both interest point detection and local description into one phase for detection. Computational efficiency of the descriptor is therefore important, leaving out many of the commonly used descriptors as unsuitably heavy. Multiresolution Gabor features has been the main descriptor in this thesis and improving their efficiency is a significant part. Actual image features are formed from descriptors by using a classifierwhich can then recognize similar looking patches in new images. The main classifier is based on Gaussian mixture models. Classifiers are used in one-class classifier configuration where there are only positive training samples without explicit background class. The local image feature detection method has been tested with two freely available face detection databases and a proprietary license plate database. The localization performance was very good in these experiments. Other applications applying the same under-lying techniques are also presented, including object categorization and fault detection.

Similarity measure based classification

Luokittelujärjestelmää suunniteltaessa tarkoituksena on rakentaa systeemi, joka pystyy ratkaisemaan mahdollisimman tarkasti tutkittavan ongelma-alueen. Hahmontunnistuksessa tunnistusjärjestelmän ydin on luokitin. Luokittelun sovellusaluekenttä on varsin laaja. Luokitinta tarvitaan mm. hahmontunnistusjärjestelmissä, joista kuvankäsittely toimii hyvänä esimerkkinä. Myös lääketieteen parissa tarkkaa luokittelua tarvitaan paljon. Esimerkiksi potilaan oireiden diagnosointiin tarvitaan luokitin, joka pystyy mittaustuloksista päättelemään mahdollisimman tarkasti, onko potilaalla kyseinen oire vai ei. Väitöskirjassa on tehty similaarisuusmittoihin perustuva luokitin ja sen toimintaa on tarkasteltu mm. lääketieteen paristatulevilla data-aineistoilla, joissa luokittelutehtävänä on tunnistaa potilaan oireen laatu. Väitöskirjassa esitetyn luokittimen etuna on sen yksinkertainen rakenne, josta johtuen se on helppo tehdä sekä ymmärtää. Toinen etu on luokittimentarkkuus. Luokitin saadaan luokittelemaan useita eri ongelmia hyvin tarkasti. Tämä on tärkeää varsinkin lääketieteen parissa, missä jo pieni tarkkuuden parannus luokittelutuloksessa on erittäin tärkeää. Väitöskirjassa ontutkittu useita eri mittoja, joilla voidaan mitata samankaltaisuutta. Mitoille löytyy myös useita parametreja, joille voidaan etsiä juuri kyseiseen luokitteluongelmaan sopivat arvot. Tämä parametrien optimointi ongelma-alueeseen sopivaksi voidaan suorittaa mm. evoluutionääri- algoritmeja käyttäen. Kyseisessä työssä tähän on käytetty geneettistä algoritmia ja differentiaali-evoluutioalgoritmia. Luokittimen etuna on sen joustavuus. Ongelma-alueelle on helppo vaihtaa similaarisuusmitta, jos kyseinen mitta ei ole sopiva tutkittavaan ongelma-alueeseen. Myös eri mittojen parametrien optimointi voi parantaa tuloksia huomattavasti. Kun käytetään eri esikäsittelymenetelmiä ennen luokittelua, tuloksia pystytään parantamaan.

Weierstrass integrability of differential equations

The integrability problem consists in finding the class of functions a first integral of a given planar polynomial differential system must belong to. We recall the characterization of systems which admit an elementary or Liouvillian first integral. We define {\it Weierstrass integrability} and we determine which Weierstrass integrable systems are Liouvillian integrable. Inside this new class of integrable systems there are non--Liouvillian integrable systems.

Cyclicity versus center problem

We prove that there are one-parameter families of planar differential equations for which the center problem has a trivial solution and on the other hand the cyclicity of the weak focus is arbitrarily high. We illustrate this phenomenon in several examples for which this cyclicity is computed.

Generalization of Vélu’s formulae for isogenies between elliptic curves

Given an elliptic curve E and a finite subgroup G, V ́lu’s formulae concern to a separable isogeny IG : E → E ′ with kernel G. In particular, for a point P ∈ E these formulae express the first elementary symmetric polynomial on the abscissas of the points in the set P + G as the difference between the abscissa of IG (P ) and the first elementary symmetric polynomial on the abscissas of the nontrivial points of the kernel G. On the other hand, they express Weierstraß coefficients of E ′ as polynomials in the coefficients of E and two additional parameters: w0 = t and w1 = w. We generalize this by defining parameters wn for all n ≥ 0 and giving analogous formulae for all the elementary symmetric polynomials and the power sums on the abscissas of the points in P +G. Simultaneously, we obtain an efficient way of performing computations concerning the isogeny when G is a rational group.

Humeral development from neonatal period to skeletal maturity—application in age and sex assessment

The goal of the present study is to examine cross-sectional information on the growth of the humerus based on the analysis of four measurements, namely, diaphyseal length, transversal diameter of the proximal (metaphyseal) end of the shaft, epicondylar breadth and vertical diameter of the head. This analysis was performed in 181 individuals (90 ♂ and 91 ♀) ranging from birth to 25 years of age and belonging to three documented Western European skeletal collections (Coimbra, Lisbon and St. Bride). After testing the homogeneity of the sample, the existence of sexual differences (Student"s t- and Mann-Whitney U-test) and the growth of the variables (polynomial regression) were evaluated. The results showed the presence of sexual differences in epicondylar breadth above 20 years of age and vertical diameter of the head from 15 years of age, thus indicating that these two variables may be of use in determining sex from that age onward. The growth pattern of the variables showed a continuous increase and followed first- and second-degree polynomials. However, growth of the transversal diameter of the proximal end of the shaft followed a fourth-degree polynomial. Strong correlation coefficients were identified between humeral size and age for each of the four metric variables. These results indicate that any of the humeral measurements studied herein is likely to serve as a useful means of estimating sub-adult age in forensic samples.

Intelligent Internet Data Searching

In this thesis author approaches the problem of automated text classification, which is one of basic tasks for building Intelligent Internet Search Agent. The work discusses various approaches to solving sub-problems of automated text classification, such as feature extraction and machine learning on text sources. Author also describes her own multiword approach to feature extraction and pres-ents the results of testing this approach using linear discriminant analysis based classifier, and classifier combining unsupervised learning for etalon extraction with supervised learning using common backpropagation algorithm for multilevel perceptron.