Local Outlier Selection Algorithm Based on Average Outlier Factor

Aiming at the shortage of subjectivity existed in the traditional local outlier detection algorithm, we studied the relevant algorithm of local outlier detection, and a local outlier selection algorithm based on the average outlier factor was proposed. This algorithm calculates the neighborhood of each data object, and figures out the outlier factor according to the dataset of its neighborhood. Comparing the outlier factor of each object with the average outlier factor to determine whether the object is an outlier. Theoretic analysis and experimental results show that this algorithm can avoid the subjectivity of selection and improve the accuracy of selection.