Explain data by explaining models
Just doing supervised learning is not enough for data minning. In real life, we maybe able to apply supervised machine learning for tasks such as inventory and sales forcasting. However, there are so much more information that people needs to manage inventory and sales. For starters, people cares more about how to increase sales and decease inventory, as far as what I have experienced?
So, it is more beneficial to understand the factors that drives the inventory up or down. Here I am trying something to extent the power of the well explored supervised machine learning algorithms to explain how factors affects inventory as such.