Machine learning and financial big data control using IoT
Jian Xiao
DOI: https://doi.org/10.3233/idt-230156
2023-09-06
Intelligent Decision Technologies
Abstract:Machine learning algorithms have been widely used in risk prediction management systems for financial data. Early warning and control of financial risks are important areas of corporate investment decision-making, which can effectively reduce investment risks and ensure companies' stable development. With the development of the Internet of Things, enterprises' financial information is obtained through various intelligent devices in the enterprise financial system. Big data provides high-quality services for the economy and society in the high-tech era of information. However, the amount of financial data is large, complex and variable, so the analysis of financial data has huge difficulties, and with the in-depth application of machine learning algorithms, its shortcomings are gradually exposed. To this end, this paper collects the financial data of a listed group from 2005 to 2020, and conducts data preprocessing and Feature selection, including removing missing values, Outlier and unrelated items. Next, these data are divided into a training set and a testing set, where the training set data is used for model training and the testing set data is used to evaluate the performance of the model. Three methods are used to build and compare data control models, which are based on machine learning algorithm, based on deep learning network and the model based on artificial intelligence and Big data technology proposed in this paper. In terms of risk event prediction comparison, this paper selects two indicators to measure the performance of the model: accuracy and Mean squared error (MSE). Accuracy reflects the predictive ability of the model, which is the proportion of all correctly predicted samples to the total sample size. Mean squared error is used to evaluate the accuracy and error of the model, that is, the square of the Average absolute deviation between the predicted value and the true value. In this paper, the prediction results of the three methods are compared with the actual values, and their accuracy and Mean squared error are obtained and compared. The experimental results show that the model based on artificial intelligence and Big data technology proposed in this paper has higher accuracy and smaller Mean squared error than the other two models, and can achieve 90% accuracy in risk event prediction, which proves that it has higher ability in controlling financial data risk.