Software Intensity Function Prediction by Haar Wavelet Regression

This paper proposes a semi-parametric model to predict the software intensity function of NHPP-based SRM. Haar wavelet is used to extract the features of the software intensity function from the observed software fault count data, and a simple quadratic function is used to predict the trend of the Haar coefficients. The prediction of the software intensity function is achieved by applying inverse Haar wavelet transform to the predicted Haar coefficients.