Optimal Subspace Analysis Based on Information-Entropy Increment

Abstract

The data structure is becoming more and more complex, and the scale of the data set is getting larger and larger. The strong limitations and instability in the high-dimensional data environment is showed in traditional outlier detection method. To solve the problems, an Optimal subspace Analysis based on Information-entropy Increment is proposed. The concepts such as mutual information and dimensional entropy to re-portrait the indicators that measure the pros and cons of subspace clustering, optimize the objective function of the clustering subspace, and obtain the optimal subspace. According to the idea of dividing the information entropy increment by one, the entropy outlier score is proposed as a metric to detect outliers in the optimal subspace. Finally, experiments verify the effectiveness of the algorithm.