Research On Big Data Generation Algorithm For Portfolio Investment Forecasting

Big data is one of the words we know well in recent years.Government departments use big data to solve resource scheduling and management problems brought about by public emergencies,the industrial sector uses big data technology to enhance product competitiveness,and the financial industry uses big data to predict changes in financial markets.Everyone is in the era of big data.However,when researcher studying big data and its processing methods,they will encounter an inevitable question where to get big data samples.And big data often carries industry secrets and is not easy to disclose,making it more difficult to obtain the required big data.So generating simulation big data integration is a valuable research topic in the IT field.Aiming at the application of portfolio investment forecasting,this paper studies a method of big data generation based on trend forecasting,which includes:(1)Research on improved time series prediction generation algorithm.This article will introduce a prediction generation algorithm for multidimensional time series data sets.(2)Bayesian network model update algorithm research.Simulate the time series trend prediction data as newly generated data over time,so that the Bayesian network model trained from historical data can be updated with the addition of new data sets.(3)Research on the search algorithm of Bayesian network node sequence.Search for data node sequences and their joint probabilities that meet the requirements of trend development,and use the node sequence set and probability as the basic data set generated by big data.(4)Research on big data random generation algorithm based on trend prediction.Based on the basic data set,a weighted random method is used to generate a large data set with customized volume.The experimental results show that compared with the traditional data generation technology,the big data set generated by the big data generation method proposed in this paper can ensure the quantity,and also ensure the original data correlation,timing and other characteristics.The generated data is of high quality and has research value for portfolio investment forecasting.