Data for: Application of online multitask learning based on least squares support vector regression in the financial market
Description
These historical transaction data of financial market are downloaded from the official website of Investing, and which constitute the following stock index dataset, bond index dataset, forex index dataset, and gold index dataset, where the web address is (https://cn.investing.com/). The stock index dataset is mainly composed of 1220 historical opening index values of the four China’s stock indices, which are the Shanghai Securities Composite Index (SSEC), the SZSE Component Index (SZI), the Growth Enterprise Index (CNT), and the SSE SME Composite Index (SZSMEPI). The time period is from January. 1st, 2014 to December. 31th, 2018. The bond index dataset is mainly composed of 1219 historical opening price values of the four China’s bond indices, which are the Shanghai Securities National Bond Index (SSEBI), the Shanghai Securities Company Bond Index (SSECBI), the Shanghai Securities Enterprise Bond Index (SSEEBI), and the Shanghai Securities 5-year Term Credit Bond Inde (SSE5YCB). The time period is from January. 1st, 2015 to December. 31th, 2019. The forex index dataset mainly consists of 1043 historical data of the exchange rate between the four currencies and RMB, which are the United States Dollar to RMB (USD-CNY), the Canadian Dollar to RMB (CAD-CNY), the Euro to RMB (EUR-CNY), and the Swiss Franc to RMB (CHF-CNY). The time period is from January. 1st, 2016 to December. 31th, 2019. The Gold index dataset is mainly composed of 1213 historical opening price values of the four precious metal spots, which are the London gold (XAU), trading-delayed gold (AUTD), London silver (XAG), and trading-delayed silver (AGTD). The time period is from January. 1st, 2015 to December. 31th, 2019. Since the four data sets are all financial time series, they can be used to verify the financial time series model. At the same time, the time series in each data set have a strong correlation, so it can be used to verify the multi-task learning model.