Build a mechanism to rationalize domain-specific tests by predicting test items that are likely to be NG due to source changes by machine learning

Future development of ML TEST

Through ML, MLTEST will reduce cost of influence range analysis and eliminate personal analysis.
This is expected to improve the efficiency of evaluation in the downstream area of software development.
Furthermore, our solution prevent the risk-included code design in the middle and upstream areas.

ML TEST output image of priority prediction of evaluation items

ML TEST output image of predictions of evaluation viewpoint and target function

ML TEST operation image (downstream process)