Predict Application response time Using Regression

November 23, 2018
The Objective is predict Web execution time of 4 different request source systems (CLC, ICL, ISYS, LIMS)
Application Response Time is a key aspect of Application Performance Monitoring (APM), and is defined as the time an application takes to respond to user requests. Nowadays, it is very important to understand what all applications are working efficiently and this model is created to predict application response time based on different ways an application is approached like Request Source system, Request Action type, Calling system, Server Instance name, Remote IP Address, Scope of the request, Request receive time etc. This Experiment is performed on web execution log data for 31st Aug18 and predicted time is scored on web execution log data for 1st Sep18. BUILD AND EVALUATE MODEL: To build and evaluate the model, firstly we have change some feature type to categorical with the help of edit metadata module and incorporated 95 percentile outlier treatment using Clip Values. Then we have used Linear Regression model to predict web Execution time for 1st Sep18 data.