Abstract
Tree-based and deep learning methods can automatically generate useful features. Not only can it enhance the original feature representation, but it can also learn to generate new features. This paper develops a strategy based on Light Gradient Boosting Machine (LightGBM or LGB) and Gated Recurrent Unit (GRU) to generate features to improve the expression ability of limited features. Moreover, a SARIMA-GRU prediction model considering the weekly periodicity is introduced. First, LightGBM is used to learn features and enhance the original features representation; secondly, GRU neural network is used to generate features; finally, the result ensemble is used as the input for prediction. Moreover, the SARIMA-GRU model is constructed for predicting. The GRU prediction consequences are revised by the SARIMA model that a better prediction can be obtained. The experiment was carried out with the data collected by Ride-hailing in Chengdu, and four predicted indicators and two performance indexes are utilized to evaluate the model. The results validate that the model proposed has significant improvements in the accuracy and performance of each component.
Highlights
Tree-based and deep learning methods can automatically generate useful features
Seasonal Autoregressive Integrated Moving Average (SARIMA) is used to decompose the original data, and the time series data is separated into different components: Trend, Seasonality and Random Residuals
In the Trend chart, it can be observed that the congestion trend from Tuesday to Thursday during the week is relatively flat, and the congestion index on Monday and Friday is relatively large
Summary
Tree-based and deep learning methods can automatically generate useful features. Can it enhance the original feature representation, but it can learn to generate new features. LightGBM is used to learn features and enhance the original features representation; secondly, GRU neural network is used to generate features; the result ensemble is used as the input for prediction. With regard to the prediction technique, machine learning has the advantage of unnecessary assumptions or prior knowledge It can automatically extract useful information from the dataset, which makes up for the shortcomings of traditional methods. I n5, the LSTM-SPRVM model and the fuzzy comprehensive evaluation-based method were leveraged to predict and rank the congestion, and a traffic congestion prediction and visualization framework based on machine learning and a fuzzy comprehensive evaluation-MF-TCPV were proposed. Ref 6 considered a prediction model based on LSTM-LGB, by weighting the consequences obtained
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
