Modified Floating Search Feature Selection Based on Genetic Algorithm

Abstract

Abstract. Classification performance is adversely impacted by noisy data .Selecting features relevant to the problem is thus a critical step in classification and difficult to achieve accurate solution, especially when applied to a large data set. In this article, we propose a novel filter-based floating search technique for feature selection to select an optimal set of features for classification purposes. A genetic algorithm is utilized to increase the quality of features selected at each iteration. A criterion function is applied to choose relevant and high-quality features which can improve classification accuracy. The method is evaluated using 20 standard machine learning datasets of various sizes and complexities. Experimental results with the datasets show that the proposed method is effective and performs well in comparison with previously reported techniques.

A Framework for Blended Sub Feature Engineering for Chronic Disease Prediction Using in-Memory Computing

“…Wrapper approaches rely on a classification algorithm to evaluate feature subsets. In general, because the feature selection process is tuned for the given classification method, the wrapper technique outperforms the filter approach [11]. To assess and choose feature subsets, filter techniques employ independent criteria that are based on the general properties of the data rather than a classification algorithm.…”

“…To assess and choose feature subsets, filter techniques employ independent criteria that are based on the general properties of the data rather than a classification algorithm. Common evaluation functions are often distance, mutual information (MI), dependence, or entropy measurements computed directly from training data [11]. It employs a filter-based strategy to choose highly representative features and a wrapper-based technique to add candidate features and assess candidate subsets in order to identify the best ones.…”

“…It employs a filter-based strategy to choose highly representative features and a wrapper-based technique to add candidate features and assess candidate subsets in order to identify the best ones. Starting with an empty set, the sequential forward search (SFS) approach adds one feature subset throughout each round until a new extracted feature that maximizes the criteria function value is found, whereas the backward elimination search (SBS) approach commences with a full feature subset and removes a feature on each iteration until a preset criterion is fulfilled [11]. Simple tools like Pandas, Pyplot, and Scikit-learn are used to build the model, which is written in Pyspark (Python) and runs on an Apache Spark cluster.…”