Anita Wasilewska

Learning and detecting emergent behavior in networks of cardiac myocytes

We address the problem of specifying and detecting emergent behavior in networks of cardiac myocytes, spiral electric waves in particular, a precursor to atrial and ventricular fibrillation. To solve this problem we: (1) apply discrete mode abstraction to the cycle-linear hybrid automata (CLHA) we have recently developed for modeling the behavior of myocyte networks; (2) introduce the new concept of spatial superposition of CLHA modes; (3) develop a new spatial logic, based on spatial superposition, for specifying emergent behavior; (4) devise a new method for learning the formulae of this logic from the spatial patterns under investigation; and (5) apply bounded model checking to detect the onset of spiral waves. We have implemented our methodology as the EMERALD tool suite, a component of our EHA framework for specification, simulation, analysis, and control of excitable hybrid automata. We illustrate the effectiveness of our approach by applying EMERALD to the scalar electrical fields produced by our CELLEXCITE simulation environment for excitable-cell networks.

Bayesian network multi-classifiers for protein secondary structure prediction

Successful secondary structure predictions provide a starting point for direct tertiary structure modelling, and also can significantly improve sequence analysis and sequence-structure threading for aiding in structure and function determination. Hence the improvement of predictive accuracy of the secondary structure prediction becomes essential for future development of the whole field of protein research. In this work we present several multi-classifiers that combine the predictions of the best current classifiers available on Internet. Our results prove that combining the predictions of a set of classifiers by creating composite classifiers is a fruitful one. We have created multi-classifiers that are more accurate than any of the component classifiers. The multi-classifiers are based on Bayesian networks. They are validated with 9 different datasets. Their predictive accuracy results outperform the best secondary structure predictors by 1.21% on average. Our main contributions are: (i) we improved the best know predictive accuracy by 1.21%, (ii) our best results have been obtained with a new semi naïve Bayes approach named Pazzani-EDA and (iii) our multi-classifiers combine results of previously build classifiers predictions obtained through Internet, thanks to our development of a Java application.

Topological Rough Algebras

It is known ([15]) that the propositional aspect of rough set theory is adequately captured by the modal system S5. A Kripke model gives the approximation space (A,R) in which well formed formulas are interpreted as rough sets. Banejee and Chakraborty ([1]) introduced a new binary connective in S5, the intended interpretation of which was the notion of rough equality, defined by Pawlak in 1982. They called the resulting Lindenbaum-Tarski like algebra a rough algebra. We show here that their rough algebra is a particular case of a quasi-Boolean algebra (as introduced in [4]). It also leads to a definition of the new classes of algebras, called topological quasi-Boolean algebras2 and topological rough algebras. We introduce, following Rasiowa and Bialynicki-Birula’s representation theorem for the quasi-Boolean algebras ([4], [20]), a notion of quasi field of sets and generalize it to a new notion of a topological quasi field of sets. We use it to give the representation theorems for the topological quasi-Boolean algebras and topological rough algebras, and hence to provide a mathematical characterization of the rough algebra.

Data Mining: Foundations and Practice

This book contains valuable studies in data mining from both foundational and practical perspectives. The foundational studies of data mining may help to lay a solid foundation for data mining as a scientific discipline, while the practical studies of data mining may lead to new data mining paradigms and algorithms. The foundational studies contained in this book focus on a broad range of subjects, including conceptual framework of data mining, data preprocessing and data mining as generalization, probability theory perspective on fuzzy systems, rough set methodology on missing values, inexact multiple-grained causal complexes, complexity of the privacy problem, logical framework for template creation and information extraction, classes of association rules, pseudo statistical independence in a contingency table, and role of sample size and determinants in granularity of contingency matrix. The practical studies contained in this book cover different fields of data mining, including rule mining, classification, clustering, text mining, Web mining, data stream mining, time series analysis, privacy preservation mining, fuzzy data mining, ensemble approaches, and kernel based approaches. We believe that the works presented in this book will encourage the study of data mining as a scientific field and spark collaboration among researchers and practitioners.

Learning and Detecting Emergent Behavior in Networks of Cardiac Myocytes

We address the problem of specifying and detecting emergent behavior in networks of cardiac myocytes, spiral electric waves in particular, a precursor to atrial and ventricular fibrillation. To solve this problem we: (1) Apply discrete mode-abstraction to the cycle-linear hybrid automata ( clha ) we have recently developed for modeling the behavior of myocyte networks; (2) Introduce the new concept of spatial-superpositionof clha modes; (3) Develop a new spatial logic, based on spatial-superposition, for specifying emergent behavior; (4) Devise a new method for learning the formulae of this logic from the spatial patterns under investigation; and (5) Apply bounded model checking to detect (within milliseconds) the onset of spiral waves. We have implemented our methodology as the Emerald tool-suite, a component of our eha framework for specification, simulation, analysis and control of excitable hybrid automata. We illustrate the effectiveness of our approach by applying Emerald to the scalar electrical fields produced by our CellExcite simulator.

Mobile applications for informal economies

The proliferation of mobile phones across the world, to people of all statures, has provided platform to bring computing resources to the masses. In this paper we present three mobile phone applications designed to aid the businesses and people of informal economies in developing countries. The goal of the applications is to assist in the growth of the economy through financial education and awareness and to assist and further the literacy. Each application is motivated and inspired by the needs of the women at the Saint-Louis Womens Business Center Incubator in Senegal, Africa. We present the motivations, application design and operation, and the lessons learned throughout a year of development.

Interactive inductive learning

Abstract We propose an interactive probabilistic inductive learning model which defines a feedback relationship between the user and the learning program. We extend previously described learning algorithms to a conditional model previously described by the authors, and formulate our Conditional Probabilistic Learning Algorithm (CPLA), applying conditions as introduced by Wasilewska to a probabilistic version of the work of Wong and Wong. We propose the Condition Suggestion Algorithm (CSA) as a way to use the syntactic knowledge in the system to generalize the family of decision rules. We also examine the semantic knowledge of the system implied by the suggested conditions and analyse the effects of conditions on the system. CPLA/CSA has been implemented by the first author and was used to generate the examples presented.

Syntactic decision procedures in information systems

There are a number of algebraic models of information systems. They have been proposed by Codd (1972) , Salton (1968) , Scott (1970) and others. We deal here with a model which is the basis of a rough set investigations ( Orlowska, 1984 ; Pawlak, 1982 ; Pawlak, 1984 ). This model was proved in ( Marek, 1985 ) to be equivalent with the Codds model of relational database with one schema. We focus here on purely syntactical problems within this model. In particular we point out problems which can be solved using the automatic syntactic methods. We do it by first constructing, for a given system S its description language ℒ S . Then we define a set of a Gentzen-like ( Gentzen, 1934 ) transformation rules for its terms and describe an easy programmable procedure which generates the answers for queries submitted to the system. We show how to extend this procedure to a procedure for generating the equivalent normal form of a given term. This leads to a method of constructing not only definable sets within a given system, but also all its elementary components.

Data Preprocessing and Data Mining as Generalization

We present here an abstract model in which data preprocessing and data mining proper stages of the Data Mining process are are described as two different types of generalization. In the model the data mining and data preprocessing algorithms are defined as certain generalization operators. We use our framework to show that only three Data Mining operators: classification, clustering, and association operator are needed to express all Data Mining algorithms for classification, clustering, and association, respectively. We also are able to show formally that the generalization that occurs in the preprocessing stage is different from the generalization inherent to the data mining proper stage.

A classification model: syntax and semantics for classification

We present here Semantic and Descriptive Models for Classification as components of our Classification Model (definition [17]). We do so within a framework of a General Data Mining Model (definition [4]) which is a model for Data Mining viewed as a generalization process and sets standards for defining syntax and semantics and its relationship for any Data Mining method. In particular, we define the notion of truthfulness, or a degree of truthfulness of syntactic descriptions obtained by any classification algorithm, represented within the Semantic Classification Model by a classification operator. We use our framework to prove (theorems [1] and [3]) that for any classification operator (method, algorithm) the set of all discriminant rules that are fully true form semantically the lower approximation of the class they describe. The set of characteristic rules describes semantically its upper approximation. Similarly, the set of all discriminant rules for a given class that are partially true is semantically equivalent to approximate lower approximation of the class. The notion of the approximate lower approximation extends to any classification operator (method, algorithm) the ideas first expressed in 1986 by Wong, Ziarko, Ye [9] , and in the VPRS model of Ziarko [10].