Author(s)

Hongkai Zhao, Yijiu Zeng, Zhekai Peng

Corresponding Author

Hongkai Zhao

ABSTRACT

Taking one day as the research cycle, we define the difference between the highest value and the lowest value of the stock price in a day as the extreme difference, and the difference between the stock closing price and the opening price as the beginning and end difference. First use the time series model to construct a trend chart of trading volume and time, and roughly estimate the missing value at 2020.03.26. After that, the correlation analysis between volume and range and end-to-end difference is carried out. On this basis, the bp neural network model was established, and the specific functional relationship between the volume and the range and the end-to-end difference was constructed to predict the missing value. The missing value was compared with the time series model to estimate the missing value. It was found that the difference was not large, so I accepted This missing value. For investment strategies and investment plans, first construct a binary linear regression function of trading volume, range and end-to-end difference, and then use matlab to find its extreme value, and record the data corresponding to the range and end-to-end difference when the extreme value is obtained (x1, x2) , By calculating the unbiased estimator of the continuous compound interest return ui and its standard deviation, using the latter as the daily standard return, the index volatility in the next year can be calculated.

KEYWORDS

Range, end-to-end difference, trading volume, time series, bp neural network algorithm, binary linear regression, particle swarm algorithm, k-line graph, looking back volatility algorithm