抽象的

Analysis of multilayer perceptron machine learning approach in classifying protein secondary structures

Leo Dencelin X, Ramkumar T

Protein secondary structure prediction is an important problem in bioinformatics and transforming biomedical big data into valuable knowledge is also a quite interesting and challenging task. Various machine learning algorithms have been widely applied in bioinformatics to extract knowledge from protein data. In recent years, multilayer perceptron, emerging on the basis of artificial neural networks, is making major advances in many domains. In this paper, we have focused on analysing the machine learning based Multilayer Perceptron (MLP) algorithms in protein secondary structure prediction using different set of input features and network parameters in distributed computing environment. Overall, the MLP analysis results in classifying protein secondary structures are encouraging, the accuracy and performance are overwhelming by passing various input features and it outperforms when it will be implemented in various distributed environment. The experimental result shows that the multilayer perceptron machine learning algorithm models outperforms the other machine learning approaches.