Automatic forecasting agent for e-Commerce applications

Forecasting has been playing an important role in many industries such as marketing, financial planning, and production control. This paper firstly gives a brief review of time series forecasting techniques, particularly covers Moving Averages model and Exponential Smoothing mode. We, then applies an illustrative example to these techniques with actual data and compares four different results by introducing the concept of Mean Square Error (MSE) and Mean Absolute Error (MAE). The technique can be used in an implementation of autonomous agent, which can be used in any e-commerce as a back end computation mechanism for applications. Finally wet conclude by summarizing the results and discussing avenues for future research.