Data classification method based on Gia system

The invention provides a data classification method based on a Gia system. The method is realized based on an extreme learning machine, and the Gia system has better execution efficiency compared with other distributed calculation systems due to an updated underlying stream processing calculation architecture and various global optimization and execution optimization technologies. Massive throughput can be realized in a batch processing mode, and top-speed response can be realized in a stream processing mode; in addition, since the extreme learning machine is a single-hidden-layer feedforward neural network, the learning speed is higher than that of other learning machines; the two advantages are combined, so that data classification with higher efficiency can be realized. Furthermore, the batch processing environment or the stream processing environment is automatically judged and started through the constructed initial data set UClletion, so that the user does not need to concern the type and the source of