Demand Forecasting System BD Forecast: Bayes-Deep Learning Forecast
BD Forecast is a demand forecasting system with 2 types of algorithm. One is deep learning(machine learning) and the other one is Bayesian theorem.
Forecasting of future demands is essential to supply chain decisions and forms the basis for all optimization system of supply chain. BD Forecast is a system for forecasting of future demands that incorporates an optimization engine to estimate a time-series with some events. Our system uses a new forecasting algorithm based on robust optimization technique. The mathematical programming formulation of robust model is flexible enough to deal with many practical situations. Our model uses the concept of “event” and estimates the effect of events using the optimization model. Seasonal demands and holiday effects are easily incorporated using the events. Furthermore, our system provides optimized parameters for several time series forecasting methods such as moving average, exponential smoothing, and also Winter’s time and seasonality corrected exponential smoothing methods.