Efficient feature subset selection and classification using levy flight-based cuckoo search optimization with parallel support vector machine for the breast cancer data

The complexity of breast cancer classification is a drastic concern for a long time. Therefore, different algorithms for data mining have been developed and used to effectively classify the dataset of breast cancer. But, due to inefficient subset choice of this operation, these algorithms have problems in accuracy of classification. In this article, levy flight-based cuckoo search optimization (LFCSO) with parallel support vector machine (PSVM) is proposed for enhancing the classification accuracy of breast cancer. The proposed LFCSO-PSVM technique includes three key steps: (i) preprocessing, (ii) feature selecting subset, and (iii) classification. Preprocessing is carried out using k-means clustering that is required to eliminate the values. To acquire maximum detailed feature, these attributes are transmitted to the process the feature selecting subset, which is executed by LFCSO algorithm, and the aspiration function is used to determine the essential feature on the basis of proper fitness values. Then, the PSVM algorithm is implemented to classify the model for training and testing. This is used to increase the accuracy of the classification for the breast cancer dataset. The experimental outcomes demonstrate that the proposed LFCSO-PSVM algorithm provides higher efficiency of classification than the existing algorithms in terms of precision, recall, f-measurement, and reliability.

Automated summary

In this article, a levy flight-based cuckoo search optimization (LFCSO) with parallel support vector machine (PSVM) technique is proposed to enhance the classification accuracy of breast cancer. The experimental results demonstrate that the proposed LFCSO-PSVM algorithm provides higher efficiency of classification than the existing algorithms in terms of precision, recall, f-measurement, and reliability.