Research on the Application of Time Convolution Series in Futures Price Forecasting

Aiming at the problem of anomaly detection time series data with unbalanced distribution among classes, a method based on depth convolution neural network is proposed. With increasing trading scale futures market, it covers more and economic financial fields, volatility market intense, which constantly presents many complex phenomena that cannot be explained by other classical theories. Investors predict stock prices statistical analysis or simple machine learning methods, but because nonlinear dynamic system, these methods have huge limitations. By using existing chart obtained game master software for prediction analysis, advantages several technical indicators been effectively complemented, accuracy has improved. On basis concept attribute set measure, aiming pattern recognition, this paper proposes recognition method, establishes an mathematical model, combines gradually applied to artificial intelligence, fields. It successfully in price trend. Building forecasting model reveal inherent law implied index show its evolution mechanism can improve control ability deal risks, provide objective rigorous regulatory authorities formulate relevant policies, also significance study.