KNN Based Peak-LOF for Outlier Detection

Data mining is a concept of getting relevant data from massive databases. Nowadays, there is a lack of Prediction of disease is increased predominantly. Hence, we need to remove the inaccurate data that has been spread over in the dataset, which leads to inaccurately predicting disease. Outliers are more prone to prediction mechanisms. This paper includes the detection of outliers in disease prediction mechanisms. The goal of this paper is to bring out exact data instead of getting invalid data by finding outliers and handling them with care. Outliers can be detected using a KNN based peak-LOF based approach with manhattan distance metrics. In this paper, a proper study on outlier detection mechanisms has been done.