Time series prediction method combined with multiple models

Abstract

The invention discloses a time sequence prediction method combined with multiple models. The method comprises the following steps: decomposing a time sequence into three time sequences of long-term trend, seasonal fluctuation and irregular fluctuation by using an STL decomposition algorithm; the method comprises the following specific steps: step 1, data preprocessing; obtaining a training set, a verification set and a test set; step 2, according to the obtained time sequence of the long-term trend, adopting a traditional prediction method ARIMA model and Prophet model for prediction, and selecting a model with the best precision; step 3, according to the obtained time sequence of seasonal fluctuation, seasonal prediction is carried out by adopting an adaptive filtering method, an ARIMA model and a Prophet model, and a model with the best accuracy is selected; step 4, according to the obtained time sequence of irregular fluctuation, predicting the time sequence of irregular fluctuation by using a neural network model; and step 5, adding the prediction results from the step 2 to the step 4 to obtain a final prediction result.

Description

technical field [0001] The present invention relates to the field of traditional methods and deep learning artificial intelligence, in particular to a time series prediction method combining multiple models. Background technique [0002] Today, with the rapid development of science and technology, the operation and maintenance of production machines in many traditional enterprises has gradually shifted from manual judgment and screening to computer control and early warning. During this transformation process, traditional enterprises cannot convert the data into a method for predicting the future state of the machine, and cannot realize the real judgment and identification of the machine state in the face of massive machine state data. Therefore, we extract useful information from data that changes over time to help traditional enterprises. [0003] At present, there are many methods for time series prediction and anomaly detection, such as traditional methods, machine lear...

AI Technical Summary

Problems solved by technology

Among them, the traditional methods include ARIMA, Prophet, AR, MA, etc. These methods can only find the regularity between the data from the existing time series, and cannot accurately capture and fit the characteristics of the time series itself, etc., and cannot really improve the prediction. The accuracy; the main method of machine learning and deep learning is to convert data into supervised data sets to make predictions

Machine learning methods include linear regression, decision tree, XGBOOST, etc. Deep learning mainly includes MLP (multilayer neural network), RNN (cyclic neural network), LSTM (long short-term memory network), CNN (convolutional neural network), these methods The trend and seasonality of the time series cannot be excavated, and there is still a big gap in predicting the time series

Images