A Comprehensive Study of Machine Learning Approach on Cytological Data for Early Breast Cancer Detection

摘要

Day by day, biological data is increasing and researchers fetch various problems to extract informative data from a large scale of dataset. According to available statistical estimates 8.8 million people died every year due to cancer worldwide. Breast cancer is the second most leading cancer occurring in women. Feature selection, classification, neural network, ensemble methods spreads the research area and makes easy to find out more relevant information in the bioinformatics analysis. In our proposed model, we build a machine learning model using some efficient feature selection algorithm and then analyses it through ANN and high performance classification algorithms. We also implement ensemble methods for building our model more accurate. For verifying our model we implement cytological data of breast cancer. Chi Squared Test give the minimum relevant features where we get 99% accuracy from ANN. We use 10 fold cross validation to test the dataset.