Revenue Prediction Using Artificial Neural Network

Abstract

Predicting revenue from tenants for an enterprise having several malls cannot be easily done using conventional approach, such as spreadsheet or manual calculations. Such an enterprise has abundant data yet inadequate resources to analyze such data. This paper presents the data mining method, namely the Artificial Neural Network (ANN), to predict the revenue based on the previous data. ANN can help the enterprise by extracting the patterns formed in previous years, so that rental income can be predicted more accurately. The research was conducted based on the following phases: business and data understanding, data preparation, modeling, evaluation and deployment. Primary data were collected based on direct interviews with the management of the enterprise. The analysis was done by conducting training on the previous data to build a neural network model. Then the model is used to make predictions on rental income in subsequent years. The results showed that this model has yielded a much smaller total error value than that of previous calculation. Thus, it can be concluded that ANN can generate rental income predictions more accurate so that it can assist the enterprise in making strategic decisions based on hidden information from existing data.