ANN，GARCH下医疗器械板块预测分析

Prediction and Analysis of Medical Device Sector Based on ANN and GARCH

摘要: 2020年新型冠状病毒疫情的爆发，使得股票市场中某些板块产生剧烈的波动，尤其是医疗器械板块，本文对医疗器械板块的股票信息进行了神经网络估计与传统GARCH模型估计，发现神经网络对股票价格估计相比较于GARCH模型产生了极强的优势。基于传统计量假设的GARCH模型忽视了股市中的非线性关系，而且对于模型的假设比较苛刻，而神经网络对于原始股票序列的宽松假设，可以弥补这种缺陷。结果表明：神经网络能捕捉到医疗板块序列中的序列之间的非线性关系，在股票序列价格估计方面优于传统的GARCH模型。

Abstract: In 2020, the outbreak of New Coronavirus epidemic caused some violent fluctuations in some sectors of the stock market, especially the medical device sector. In this paper, we estimated the stock information of the medical device sector and estimated the traditional GARCH model, and found that the neural network has a strong advantage in the stock price estimation compared with the GARCH model. The GARCH model based on the traditional econometric hypothesis ignores the nonlinear relationship in the stock market, and the hypothesis of the model is harsh. The loose hypothesis of neural network for the original stock sequence can make up for this defect. The results show that: the neural network can capture the nonlinear relationship between the series in the medical plate series, and it is better than the traditional GARCH model in the estimation of stock series price.