Adaptive Data Mining Algorithm under the Massive Data

In order to solve the problem that Network Reduced accuracy and poor convergence in the existing neural network, which because sample large volumes of data and target data-independent. In response to this phenomenon, this paper put forward a data mining based on compensatory fuzzy neural network. It was optimizing was the Compensative Fuzzy Neural Network. And improve the cutting effect base on calculation algorithm. At the end, it was based on the similarity of each cluster objects to clustering process the system data. Through simulation experiments we can see, algorithm can maintain high precision under different circumstances the amount of data. Compared to other algorithms, we can see that it has a large advantage in terms of both accuracy and timeconsuming.