Clustering algorithms are important in data mining to find similarities between objects. By using a method called Auto-Encoder, data features are compressed to filter out noise. When more complex neural networks with specific functions are used, the performance of clustering methods improves in terms of both computing cost and accuracy.