Computational Analysis of MK2 Protein for HT Carcinoma Cells using Pre-Processing and Characterization Techniques for Cell Death/ Survival

Abstract

This paper presents the computational analysis using pre-processing and different characterization of MK2 proteins for cell death/ survival for HT carcinoma cells. Initially data is collected which than pre processed by completing the missing data, removing various outliers, checking whether the data is normal or non- normal. Later various characterisation steps were applied which includes: classification, regression, clustering, association rules etc. Different data variables use different approaches of pre processing and characterisation. In this paper we have collected the different data of MK2 proteins where we have cleaned the data by using various clustering approaches, than regression analysis, correlation matrix and covariance matrixes are calculated. In last we have classified the data by using different machine learning approaches (k-NN, SVM and Neural network). For k-NN we have compared the results of different techniques like chebyshev, city block and Euclidean. For SVM we have used linear, polynomial, RBF and sigmoid function for both types (Type 1 and Type 2) and for neural network we have calculated training, test and validation perfection using different hidden, and output activation function. This paper shows best 10 network using MLP and RBF approach.