Nabiha Azizi

Improved decision tree construction based on attribute selection and data sampling for fault diagnosis in rotating machines

This paper presents a new approach that avoids the over-fitting and complexity problems suffered in the construction of decision trees. Decision trees are an efficient means of building classification models, especially in industrial engineering. In their construction phase, the two main problems are choosing suitable attributes and database components. In the present work, a combination of attribute selection and data sampling is used to overcome these problems. To validate the proposed approach, several experiments are performed on 10 benchmark datasets, and the results are compared with those from classical approaches. Finally, we present an efficient application of the proposed approach in the construction of non-complex decision rules for fault diagnosis problems in rotating machines.

Optimized Tumor Breast Cancer Classification Using Combining Random Subspace and Static Classifiers Selection Paradigms

Breast cancer is the most frequently diagnosed cancer in women worldwide and the leading cause of cancer death among females. Currently the most effective method for early detection and screening of breast abnormalities is mammography. Computer aided design (CAD) systems are used to assist radiologists in better classification of tumor in a mammography as benign or malignant. Ensemble classifier construction has received considerable attention in the recent years. In the modeling of classifier ensemble, many researchers believe that the success of classifier ensembles only when classifier members present diversity among themselves. The most widely used ensemble creation techniques are focused on incorporating the concept of diversity with the construction of different features subsets or selection of the most diverse components from initial classifiers pool. Therefore the motivation of this work is to propose a CAD system using a novel classification approach based on feature selection and static classifier selection schemes.

Using diversity in classifier set selection for arabic handwritten recognition

The first observation concerning Arabian manuscript reveals the complexity of the task, especially for the used classifiers ensemble. One of the most important steps in the design of a multi-classifier system (MCS), is the its components choice (classifiers). This step is very important to the overall MCS performance since the combination of a set of identical classifiers will not outperform the individual members. To select the best classifier set from a pool of classifiers, the classifier diversity is the most important property to be considered. The aim of this paper is to study Arabic handwriting recognition using MCS optimization based on diversity measures. The first approach selects the best classifier subset from large classifiers set taking into account different diversity measures. The second one chooses among the classifier set the one with the best performance and adds it to the selected classifiers subset. The performance in our approach is calculated using three diversity measures based on correlation between errors. On two database sets using 9 different classifiers, we then test the effect of using the criterion to be optimized (diversity measures,), and fusion methods (voting, weighted voting and Behavior Knowledge Space). The experimental results presented are encouraging and open other perspectives in the classifiers selection field especially speaking for Arabic Handwritten word recognition.

From static to dynamic ensemble of classifiers selection: Application to Arabic handwritten recognition

Arabic handwriting word recognition is a challenging problem due to Arabics connected letter forms, consonantal diacritics and rich morphology. One way to improve the recognition rates classification task is to improve the accuracy of individual classifiers; another, is to apply ensemble of classifiers methods. To select the best classifier set from a pool of classifiers, the classifier diversity is considered one of the most important properties in static classifier selection. However, the advantage of dynamic ensemble selection versus static classifier selection is that used classifier set depends critically on the test pattern. In this paper, we propose two approaches for Arabic handwriting recognition AHR based on static and dynamic ensembles of classifiers selection. The first one selects statically the best set of classifiers from a pool of classifier already designed based on diversity measures. The second one represents a new algorithm based on Dynamic Ensemble of Classifiers Selection using Local Reliability measure DECS-LR. It chooses the most confident ensemble of classifiers to label each test sample dynamically. Such a level of confidence is measured by calculating the proposed local reliability measure using confusion matrixes constructed during training level. We show experimentally that both approaches provide encouraging results with the second one leading to a better recognition rate for AHR system using IFN_ENIT database.

A new hybrid method combining genetic algorithm and support vector machine classifier: Application to CAD system for mammogram images

Breast cancer continues to be one of the most common cancers, and survival rates critically depend on its detection in the initial stages. Several studies have demonstrated the benefits and potential of using CAD (Computer-Assisted Diagnosis) systems to help specialists in their clinical interpretation of mammograms. CAD is based essentially on 2 main steps: Extraction of pertinent features and classification. In fact, several types of features are used in this work characterizing the extracted masses which are: texture features based on co-occurrence matrix and shape features based on Hu moments and central moments. All these features represent feature vector used in training and testing the used classifier. To reduce dimensionality and optimize classification process a new approach based on genetic algorithm is proposed. It incorporates Svm classifier results as part of multi objective function for fitness function. Once the best subset of features is chosen, classification is made by SVM classifier using Gaussian kernel function. Experimental results demonstrate the effectiveness of the proposed algorithm.

Automated Classification of Mammographic Abnormalities Using Transductive Semi Supervised Learning Algorithm

Computer-aided diagnosis (CAD) of breast cancer is becoming a necessity given the exponential growth of performed. CAD are usually characterized by the large volume of acquired data that must be labeled in a specific way that leads to a major problem which is labeling operation. As a result the community of machine learning has attempted to respond to these practical needs by introducing the semi-supervised learning. The motivation of the current research is to propose a TSVM-CAD system for mammography abnormalities detection using a new Transductive TSVM with comparison of its kernel functions. The effectiveness of the system is examined on the Digital Database for Screening Mammography database DDSM using classification accuracy, sensitivity and specificity. Experimental results are very encouraging.

Ensemble classifier construction for Arabic handwritten recongnition

Handwritten recognition is a very active research domain that led to several works in the literature for the Latin Writing. The current systems tendency is oriented toward the classifiers combination and the integration of multiple information sources. In this paper, we describe an approach based on diversity measures for Arabic handwritten recognition using optimized Multiple classifier system. The aim of this paper is to study Arabic handwriting recognition using the optimization of MCS based on diversity measures. This approach selects the best classifier subset from a large set of classifiers taking into account different diversity measures. The experimental results presented are encouraging and open other perspectives in the domain of classifiers selection especially speaking for Arabic Handwritten word recognition.

New dynamic classifiers selection approach for handwritten recognition

In this paper a new approach based on dynamic selection of ensembles of classifiers is discussed to improve handwritten recognition system. For pattern classification, dynamic ensemble learning methods explore the use of different classifiers for different samples, therefore, may get better generalization ability than static ensemble learning methods. Our proposed DECS-LR algorithm (Dynamic Ensemble of Classifiers Selection by Local Reliability) enriched the selection criterion by incorporating a new Local-Reliability measure and chooses the most confident ensemble of classifiers to label each test sample dynamically. Confidence level is estimated by proposed reliability measure using confusion matrix constructed during training level. After validation with voting and weighted voting fusion methods, ten different classifiers and three benchmarks, we show experimentally that choosing classifiers ensemble dynamically taking into account the proposed L-Reliability measure leads to increase recognition rate for Handwritten recognition system using three benchmarks.

CAD system for classification of mammographic abnormalities using transductive semi supervised learning algorithm and heterogeneous features

Breast cancer is the most frequently diagnosed cancer in women worldwide and the leading cause of cancer death among females. Currently the most effective method for early detection and screening of breast abnormalities is mammography. In particular, the breast mass diagnosis and classification arouse nowadays a great interest. Various researches have proven that the computer-aided diagnosis (CAD) of breast abnormalities is becoming increasingly a necessity given the exponential growth of performed. Hence, it can reduce the cost for double screening process A generic CAD system includes segmentation, feature extraction, and classification stages in order to have a final decision. However, such a system is usually characterized by the large volume of the acquired data. This data must be labeled in a specific way that leads to a major problem which is the necessity of an expert to make the labeling operation. To overcome this constraint, statistical learning propose semi-supervised learning (SSL) algorithm as alternative in order to beneficiate to the all dataset images. In this paper, a CAD system for the breast abnormalities classification is proposed basing on Transductive semi-supervised learning technique using TSVM with these different kernel functions and heterogeneous features families. Experimental results based on DDSM dataset are very encouraging.

Classifier Ensemble Selection Based on mRMR Algorithm and Diversity Measures: An Application of Medical Data Classification

Classifier selection is a significant problem in machine learning to reduce the computational time and the number of ensemble members. Over the past decade, multiple classifier systems (MCS) have been actively exploited to enhance the classification accuracy. Finding a pertinent objective function for measuring the competence of base classifier is a critical issue to select the appropriate subset from a pool of classifiers. Along with the accuracy, diversity measures are designed as objective functions for ensemble selection. This current work proposed a new selection method based on accuracy and diversity in order to achieve better medical data classification performance. The classifiers correlation was calculated using Minimum Redundancy Maximum Relevance (mRMR) method based on relevance and diversity measures. Experiments were carried out on five data sets from UCI Machine Learning Repository and LudmilaKuncheva Collection. The experimental results proved the superiority of the proposed classifiers selection method.