COVID-19 PREDICTION BASED ON INVESTMENT PSYCHOLOGY AND EMOTION REGULATION UNDER NEW CROWN PNEUMONIA — APPLICATION OF SVR MODEL IN AGRICULTURAL LISTED COMPANIES

BACKGROUND: At the end of 2019, the epidemic of coronavirus disease had a negative impact on Residents' income and life. Intermittent production and shutdown measures had a negative impact on production, especially agriculture. In order to alleviate the uncertain impact of the COVID-19 in 2019 and solve the contradiction of the shortage of agricultural production and operation funds, stock listing financing has become a feasible path. The trend of the stock market reflects the current economic situation to a certain extent. The prediction of the stock market is conducive to the timely adjustment of macroeconomic policies and the maintenance of market stability. As an important part of listed companies, agricultural listed companies rely more on the accurate prediction of agricultural enterprises. Compared with other types of enterprises, agricultural enterprises may face greater uncertainty and higher risks. This paper improves the prediction accuracy of the model on the basis of quantum optimization fruit fly algorithm, quantum optimization bee colony algorithm, quantum optimization particle algorithm and quantum optimization ant colony algorithm. Taking the agricultural listed companies in the Shanghai stock index as the experimental sample, the effectiveness of quantum particle swarm optimization SVR model in the stock prediction of agricultural listed companies is verified, and sufficient comparative experiments are carried out to explore the impact of different structures on performance, until a reasonable optimization path is given to predict the stocks of listed companies, especially the investment psychology and emotional regulation factors of investors to listed companies. RESEARCH OBJECTS AND METHODS: The data in this paper are based on the stock trading of Huaying agriculture and agricultural development seed industry for 500 days, that is, the inventory change from November 20, 2018 to December 9, 2020. Through the prediction of agricultural inventory, four models (QPSO quantum particle swarm, qaco quantum ant colony, qabc quantum bee colony and qfoa quantum fruit fly) can indirectly judge the development trend and future prospect of agriculture according to the degree of fitting. These four models are used to replace cross validation, and these indicators are used to predict stock prices. In order to better study the changes of risk cognition and action, this study adopts the quantum ant and quantum fruit fly emotional change scale, which is compiled under the guidance of the theory of R.S. Weiss (1973). The purpose is to distinguish between emotional isolation (lack of intimate contact with another) and social isolation (lack of communication with friends with common interests). The two scales are conceptually interrelated, and the latter (Wittenberg 1986) is an extension of the former and has not been published. Russell et al. (1984) has only two items in the scale, which are aimed at two different kinds of loneliness. Each item had two sentences describing a way of loneliness. Subjects were asked to rate their current feeling intensity on a 9-level scale (two levels were “none at all” and “extremely heavy”). Wittenbers et al. (1986) has a total of 10 items, with 5 items each to assess social and emotional loneliness. The item score is divided into five levels, and the total score of the two sub tables is 5-25. A high score is a heavy loneliness. Russell's internal consistency: there is only one item in each subscale of Russell's scale, so there is no consistency. Wittenberg scale emotional loneliness subscale α the coefficient is 0.78, which is the sub table of social loneliness α The coefficient is 0.76. RESULTS: Taking Shanghai index agricultural listed companies as an example, the effectiveness of quantum particle swarm optimization SVR model in stock prediction of agricultural listed companies was verified. A full comparative experiment is carried out to discuss the impact of different structures on performance, until a reasonable optimization path is given to predict the stocks of listed companies, especially the stocks of agricultural listed companies. The research shows that among the QPSO model, qaco model, qabc model and qfoa model, quantum particle swarm optimization algorithm has the best prediction effect, faster error convergence and stronger stability. Quantum particle swarm optimization is also one of the most comprehensive indicators. CONCLUSION: This method is suitable for stock price prediction and has good prediction effect. However, because it can not be compared with other stock price prediction methods, it is impossible to judge this model and other models. However, this research method can also provide different ideas for other scholars who study stock price prediction. ACKNOWLEDGEMENTS: Supported by a project grant from Hunan Natural Science Foundation (No: 2021JJ30069, 2021JJ30295) and Hunan Philosophy and Social Science Foundation Project (No: 20YBA121).