Supply Chain Demand forecasting with Python
Learn to build advanced time series demand forecasting models in python
What you will learn
Timeseries data cleaning and preparation
Implement simple moving average forecast in python
Learn different KPIs (Bias, MAPE, MAE, RMSE) to measure forecast accuracy & implement in Python
Implement weighted moving average forecast & optimize parameters in python
Implement single exponential smoothing model in python
Implement double exponential smoothing model in python
Implement double exponential smoothing with damped trend
Implement triple exponential smoothing model in python
Simulate and optimize Alpha, Beta, Gamma and Phi parameters for automatic model selection
Calculate error KPIs for each models
Visualize the results with actuals, forecast and forecast errors
Why take this course?
Understanding and predicting the demand is one of the key challenge in Supply chain planning. Having better forecasting meaning better supply planning and optimized business operations with good customer service, therefore learn to build better forecast is a key skill to master in Supply chain management. Demand forecasting sounds simple but it will get complex when we have thousands of SKUs and each with its own demand pattern such as seasonal, intermittent and lumpy.
In this course you will learn demand forecasting models from basic to more advanced. And implement each of the models in Python. You will gain practical knowledge with real life data with over 3000 skus and over 5 years of data and millions of transactions.
By the time you complete the course you would have learned how advanced demand forecasting engine works in expensive commercial software and you would build your own fully automated forecasting engine.
In this course you will not only learn to build forecasting models and predict demand but also learn to build a python tool which can automatically optimize and select the best forecasting model based on your data.
Last but not least, you will learn to visualize all the forecasted data and errors in an intutive way.