QSPR Outlier Detection Methods

The high performance of QSPR may get affected when. the dataset contains outlier data which leads to misleading output. Occurrences of outlier in dataset are caused by various reasons like human error, change in the method of collecting data and data processing methods. When generating QSPR model for high dimensional data, there may be possibility of getting fault result, if outliers are not detected and handled before analysis. There occurs the situation where outlier detection and removal of outlier to be done as pre-processing step to develop a better QSPR prediction model.