A Web-based Real-time Sales Data Analysis using LSTM Model for Better Insight

The main goal of this paper is to perform a time series analysis on real-time sales data using the Long Short Term Memory Algorithm (LSTM). The investigation is performed on the various aspects of business and sales historical data to analyze how it impacts profitability, by reducing the repetitive tasks in day-to-day life. A machine learning-based web application is developed and deployed in real-time for the Nellai Agencies company to analyze their sales data. Initially, the LSTM model is trained with monthly sales data to predict the sales for the next ‘n’ days. The major steps involved are data preprocessing, selection of target feature, scalar transformations, fitting the model, and finally the prediction. The performance of the model is evaluated with the metrics such as Mean Absolute Error (MAE) & R²Score. Thus, the results obtained are used to make insights by the company by visualizing various graphs. Hence, the impact of using a machine learning-based web application supports better decision-making for the distribution of sales & production and finding the most prominent customers to increase the profitability of the company.