Chinese energy security index price forecasting through the neural network

For investors and governments, predicting the price of energy security is a critical problem. This research examines the suitability of the nonlinear autoregressive neural network for this forecasting problem using a dataset of daily closing prices for the energy security index that was traded on the China Shanghai Stock Exchange between January 4, 2016, and December 31, 2020. To create a model that performs accurately and reliably, a variety of model settings for the algorithm, delay, hidden neuron, and data splitting ratio was investigated. Energy security pricing forecasting is shown to benefit greatly from machine learning technologies. The findings may be paired with basic estimations to perform policy research and provide opinions on price patterns, or they could be used alone as technical forecasts.