Abstract:

Detecting outliers in the data set is quite important for building effective predictive models. Consistent prediction can not be made through models created with data sets containing outliers, or robust models can not be created. In such cases, it may be possible to exclude observations that are determined to be outlier from the data set, or to assign less weight to these points of observation than to other points of observation. Lower and upper boundaries can be created to exclude outliers from the dataset, and models can be created using the data between those boundaries. In this study, it was aimed to propose a different perspective on outlier detection methods by creating upper bounds with the aid of deep neural networks using skewness, kurtosis and standard deviation values obtained from the dataset with trained models.  

Keywords: