S Sabeena ¹, B. Sarojini ¹

Affiliations
1 Department of Computer Science, Avinashilingam Institute for Home Science and Higher Education for Women, Coimbatore - 641043, Tamil Nadu, IN

Abstract

Background/Objectives: Data mining is the process of extracting large volumes of raw data from hidden knowledge. The health care industry requires the use of data mining techniques as it generates huge and complex volumes of data. The applications of data mining techniques to medical data extract patterns which are useful for diagnosis, prognoses and treatment of diseases. This extraction of patterns allows doctors and hospitals to be more effective and more efficient. The huge volume of data is the barrier in the detection of patterns. Feature selection techniques mainly used in data preprocessing for data mining. Methods/Statistical Analysis: Classification task leads to reduction of the dimensionality of feature space, feature selection process is used for selecting large set of features. The Ant Colony Optimization based feature selection method is applied on cancer datasets. Findings: This research work proposes about feature selection mechanism based on Ant Colony Optimization. In an ACO algorithm, the activities of ants have significance for solving different combinatorial optimization problem which selects most relevant features. Through several iterations filter based method finds the optimal feature subset. Based on the similarity between features the feature relevance will be computed, that shows to the minimization of the redundancy. To validate the proposed feature selection method Support Vector Machine classification is applied. The accuracy of classification for whole feature set and the reduced feature subset are compared. The improved accuracy proves that the proposed feature selection approach has selected informative feature of the cancer datasets. Applications/Improvements: The possibilities of using PSO algorithm is applied for finding the best features in future. Other algorithms are also considered for further implementation.

Keywords

Ant Colony Optimization, Feature Selection, Support Vector Machine

Full Text

   

G. Priyadharshini ¹, B. Sarojini ¹

Affiliations
1 Department of Computer Science, Avinashilingam Institute for Home Science and Higher Education for Women, Coimbatore - 641043, Tamil Nadu, IN

Abstract

Background/Objectives: Feature selection is one of the preprocessing techniques used for removing redundant and irrelevant features. The objective of this research work is to show that small set of relevant features can improve the performance of classification algorithms. Methods/Statistical Analysis: This paper compares the performance of two feature selection algorithms, Modified Mutual Information based Feature Selection (MMIFS) and Dynamic Mutual Information based Feature Selection (DMIFS) The performance of these feature selection algorithms on the medical datasets is analyzed. The performances of c4.5 classification algorithm before and after feature selection are analyzed. Findings: The comparative study show that the feature selection algorithms have selected prominent features of the medical datasets. The percentage of feature reduction and the improvement in the accuracy of the classification algorithm are used for validation. The result shows an improvement in the accuracy of the classification algorithm. Applications/ Improvements: The reduction in the number of features means diagnosis of the disease with limited number of relevant features. Integrating feature selection techniques and machine learning algorithms will give a better decision making tool which is appreciable in medical domain.

Keywords

Decision Tree, Dynamic Mutual Information based Feature Selection, Feature Selection, Modified Mutual Information based Feature Selection

Full Text

   

Bhuvaneswari Ragothaman ¹, B. Sarojini ¹

Affiliations
1 Department of Computer Science, Avinashilingam Institute for Home Science and Higher Education for Women - University, Coimbatore - 641043, Tamil Nadu, IN

Abstract

Objective: There exists a huge amount of heterogeneous data in medical databases which when mined may provide valuable information for medical diagnosis. The processing of this voluminous data is tedious, owing to its high dimensionality. The presence of irrelevant and redundant data may reduce the performance of data mining algorithms. The data preprocessing technique, feature selection, removes any irrelevant or redundant features and selects discriminative features. The objective of this research work is to propose a feature selection algorithm that meets two objectives, 1. Reduce the number of features and 2. Maximize the classification accuracy. Methods/Statistical Analysis: The research work proposes a feature selection algorithm Multi-Objective Non-Dominated Sorted Artificial Bee Colony Algorithm (NSABC) that combines Pareto optimization and Artificial Bee Colony (ABC) algorithm for selecting the Non-Dominant optimal feature subsets of three medical datasets viz. Wisconsin Breast Cancer, Pima Indian Diabetes and Statlog Heart disease datasets. The features selected by the ABC algorithm are further optimized by applying Pareto optimization. The feature subsets selected are validated by evaluating the performance of KNN classifier in terms of the classification metrics Precision, Recall and Accuracy before and after feature selection. Findings: The percentage of feature reduction and the improved performance of the classifier prove that the proposed feature selection method has selected the discriminate features of the datasets. The selected feature subsets are further validated by calculating the entropy, the measure signifying the individuality and independent nature of the features in the feature subset. Application/Improvement: The reduced feature subset selected is used in the easy diagnosis of the disease during the medical diagnosis. This also helps in cost and time reduction for the medical diagnosis of the disease.

Keywords

Data Mining, Entropy Feature Selection, KNN Classifier, Non-Dominated Sorted Artificial Bee Colony Algorithm (NSABC), Pareto Optimization.

Full Text