ANN-PSO Approach on Logistic Model based Software Reliability Prediction
Manjubala Bisi · 2020
Reliable software can be produced within given time and budget constraints using software reliability prediction model. In this paper, Logistic Growth Curve Model (LGCM) based on ANN-PSO approach to predict software reliability is proposed. Different types of activation functions are considered in each of the hidden layer of the Artificial Neural Network (ANN). Logistic growth curve model is represented using an ANN to predict software reliability. An ANN-PSO approach based LGCM is proposed to optimize the weights of ANN using PSO. Three data sets have been considered for validating the proposed model. DWCM model, LGCM-NN and LGCM-GA models are considered for comparison with the proposed approach. The proposed model shows better efficacy than the compared models.