Lawrence J. Mazlack

Identifying potential to acquire programming skill

Over several years, a large number of students took the same introductory programming course. The students were drawn from all academic disciplines and academic experience levels. Low correlations were found between success in the course or its evaluation components when posed against academic program, gender, or semester in school. Additionally, when the IBM Programmers Aptitude Test was administered, the predictive value of the test was found to be low.

Fuzzy-rough nearest-neighbor classification approach

This paper proposes a new fuzzy-rough nearest-neighbor (NN) approach based on the fuzzy-rough sets theory. This approach is more suitable to be used under partially exposed and unbalanced data set compared with crisp NN and fuzzy NN approach. Then the new method is applied to China listed company financial distress prediction, a typical classification task under partially exposed and unbalanced learning space. Results suggest that the compared with crisp and fuzzy nearest neighbor classification methods, this method provides more accurate prediction result under this research design.

Learning fuzzy cognitive maps from data by ant colony optimization

Fuzzy Cognitive Maps (FCMs) are a flexible modeling technique with the goal of modeling causal relationships. Traditionally FCMs are developed by experts. We need to learn FCMs directly from data when expert knowledge is not available. The FCM learning problem can be described as the minimization of the difference between the desired response of the system and the estimated response of the learned FCM model. Learning FCMs from data can be a difficult task because of the large number of candidate FCMs. A FCM learning algorithm based on Ant Colony Optimization (ACO) is presented in order to learn FCM models from multiple observed response sequences. Experiments on simulated data suggest that the proposed ACO based FCM learning algorithm is capable of learning FCM with at least 40 nodes. The performance of the algorithm was tested on both single response sequence and multiple response sequences. The test results are compared to several algorithms, such as genetic algorithms and nonlinear Hebbian learning rule based algorithms. The performance of the ACO algorithm is better than these algorithms in several different experiment scenarios in terms of model errors, sensitivities and specificities. The effect of number of response sequences and number of nodes is discussed.

Detecting brain structural changes as biomarker from magnetic resonance images using a local feature based SVM approach

Detecting brain structural changes from magnetic resonance (MR) images can facilitate early diagnosis and treatment of neurological and psychiatric diseases. Many existing methods require an accurate deformation registration, which is difficult to achieve and therefore prevents them from obtaining high accuracy. We develop a novel local feature based support vector machine (SVM) approach to detect brain structural changes as potential biomarkers. This approach does not require deformation registration and thus is less influenced by artifacts such as image distortion. We represent the anatomical structures based on scale invariant feature transform (SIFT). Likelihood scores calculated using feature-based morphometry is used as the criterion to categorize image features into three classes (healthy, patient and noise). Regional SVMs are trained to classify the three types of image features in different brain regions. Only healthy and patient features are used to predict the disease status of new brain images. An ensemble classifier is built from the regional SVMs to obtain better prediction accuracy. We apply this approach to 3D MR images of Alzheimers disease, Parkinsons disease and bipolar disorder. The classification accuracy ranges between 70% and 87%. The highly predictive disease-related regions, which represent significant anatomical differences between the healthy and diseased, are shown in heat maps. The common and disease-specific brain regions are identified by comparing the highly predictive regions in each disease. All of the top-ranked regions are supported by literature. Thus, this approach will be a promising tool for assisting automatic diagnosis and advancing mechanism studies of neurological and psychiatric diseases.

Inferring Fuzzy Cognitive Map models for Gene Regulatory Networks from gene expression data

Gene Regulatory Networks (GRNs) represent the causal relations among the genes and provide insight on the cellular functions and the mechanism of the diseases. GRNs can be inferred from gene expression data by a number of algorithms, e.g. Boolean networks, Bayesian networks, and differential equations. While reliable inference of GRNs is still an open problem, new algorithms need to be developed. Fuzzy Cognitive Maps (FCMs) is used to represent GRNs in this paper. Most of the FCM learning algorithms are able to learn FCMs with less than 40 nodes. A new algorithm that is able to learn FCMs with more than 100 nodes is proposed. The proposed method is based on Ant Colony Optimization (ACO). A decomposed approach is proposed to reduce the dimension of the problem; therefore the FCM learning algorithm is more scalable (the dimension of the problem to be solved in one ACO run equals to the number of nodes or genes). The proposed approach is tested on data from DREAM project. The experiment results suggest the proposed approach outperforms several other algorithms.

Inferring causal networks using fuzzy cognitive maps and evolutionary algorithms with application to gene regulatory network reconstruction

Proposed a decomposed evolutionary algorithm for learning fuzzy cognitive map models.Applied the proposed method to infer causal networks from gene expression time series.Learned 300-node fuzzy cognitive maps while less than 40 nodes are common in the literature.Compared four stochastic optimization methods for learning fuzzy cognitive maps. Fuzzy cognitive maps have been widely used as abstract models for complex networks. Traditional ways to construct fuzzy cognitive maps rely on domain knowledge. In this paper, we propose to use fuzzy cognitive map learning algorithms to discover domain knowledge in the form of causal networks from data. More specifically, we propose to infer gene regulatory networks from gene expression data. Furthermore, a new efficient fuzzy cognitive map learning algorithm based on a decomposed genetic algorithm is developed to learn large scale networks. In the proposed algorithm, the simulation error is used as the objective function, while the model error is expected to be minimized. Experiments are performed to explore the feasibility of this approach. The high accuracy of the generated models and the approximate correlation between simulation errors and model errors suggest that it is possible to discover causal networks using fuzzy cognitive map learning. We also compared the proposed algorithm with ant colony optimization, differential evolution, and particle swarm optimization in a decomposed framework. Comparison results reveal the advantage of the decomposed genetic algorithm on datasets with small data volumes, large network scales, or the presence of noise.

Granular causality speculations

In many ways, causality deals with granular descriptions. This is true for commonsense reasoning as well as for mathematical and scientific theory. At a very fine-grained level, the physical world itself may be made up out of granules. Our commonsense perception of causality is often granular. Knowledge of at least some causal effects is imprecise. In our commonsense world, it is unlikely that all possible factors can be known. Our commonsense understanding of the world deals with imprecision, uncertainty and imperfect knowledge. People recognize that a complex collection of elements causes a particular effect, even if the precise elements of the complex are unknown. They may not know what events are in the complex; or, what constraints and laws the complex is subject to. Sometimes, the details underlying an event can be known to a fine level of detail, sometimes not. The level of detail can reasonably be called the events grain size. In general, commonsense reasoning is more successful in reasoning about a few large grain sized events than many fine-grained events. The central question is: to what extent can we usefully vary the causal grain size?.

Computer construction of crossword puzzles using precedence relationships

This paper reports on the construction of a crossword puzzle generator. After an unsuccessful attempt to construct puzzles by whole insertion, puzzles were successfully constructed by a letter by letter method. A dynamic, heuristically determined, decision structure was required. The constructor resolved questions of letter selection, ordering and reordering of the solution sequence, dictionary structure and access, and decision path selection.

Representing Causality Using Fuzzy Cognitive Maps

Causal reasoning occupies a central position in human reasoning. In order to algorithmically consider causal relations, the relations must be placed into a representation that supports manipulation. The most widespread causal representation is directed acyclic graphs (DAGs). However, DAGs are severely limited in what portion of the common sense world they can represent. This paper considers the needs of commonsense causality and suggests Fuzzy Cognitive Maps as an alternative to DAGs.

Machine Selection of Elements in Crossword Puzzles: An Application of Computational Linguistics

This paper reports on the construction of a crossword puzzle generator. After an unsuccessful attempt to construct puzzles by whole word insertion, puzzles were constructed letter by letter. Heuristically determined decision structure was required. The constructor resolved questions of letter selection, ordering and reordering of the solution sequence, dictionary structure and access, and decision path selection. The decision basis for letter selection was based on a pseudo-probabilistic approach.