标题:
基于分类预测的启发式错误定位方法Heuristic Wrong Positioning Method Based on Classification Forecast
作者:
王瑾, 罗杰
关键字:
机器学习方法, 分类预测, 启发式, 错误定位Machine Learn Method, Classification Forecast, Heuristic, Wrong Positioning
期刊名称:
《Software Engineering and Applications》, Vol.5 No.1, 2016-02-17
摘要:
本文利用机器学习的方法通过训练学习,对通过形式化的错误定位算法进行错误定位的程序在二次定位或者多次定位前进行结果预测,引导用户给出高效的反馈信息(期望值或者正确值的确认信息),从而减少错误定位的次数,提高错误定位的效率,实现启发式的错位定位。
This paper mainly forecasts the wrong positioning results for the program of the formal error localization algorithm before using it proceed two wrong positioning or multiple wrong positioning by training learn based on machine learn methods. It leads the user to give the high effective feedback information (expected value or the right value information). Thus, it decreases the wrong positioning number, raises the efficiency of the wrong positioning and realizes the heuristic wrong positioning.