Research on RNA Secondary Structure Prediction Based on MLP.

RNA has important structure, function and regulation in cells. It is closely related to biological functions. Pseudoknot is an important secondary structure in RNA. It is the most difficult part to predict. In order to solve this problem, this article attempts to use Multi-Layer Perceptron (MLP) neural network to predict and study RNA secondary structure. And compared with support vector machine (SVM), random forest (Random Forests), decision tree (Decision Tree) and other algorithms on the authoritative data set RNA STRAND. Experiments show that the classification accuracy and model score of MLP are more advantageous, which verifies the feasibility and effectiveness of the MLP method.