Analysis and Prediction of Business and Financial Data Based on LSTM Neural Network

In the context of today’s informatization and big data, the traditional financial management model of power grid projects has fallen behind. The financial department is unable to calculate and control key indicators in the process of engineering projects, but can only calculate the results and solidify assets. In response to such problems, in order to advance the role of financial to important nodes in the process of projects, this paper proposes a method for analyzing and predicting business financial data based on LSTM neural network, which uses multi-source heterogeneous data at the current stage of the project to affect future costs Forecast with income, and the financial department calculates indicators based on the forecast results. When the indicators are poor, the project decision of the current stage can be adjusted to realize the closed-loop monitoring of the business process. The results show that the method is accurate and effective.