Machine Learning

In the talk i also gave a detailed recipe of what we did for forecasting at the scale of shopping five, but i can sheeralilit some insighe on my wat what i would recommand there. The way i try to think about it is i always step back and say, what's the problem that shopofi needs me to solve at this minute? And then try to really understand what good looks like. Don't go from a if then else condition to a deep learning like, don't do that jump. Start simple and iterate overtime, and then apply more sophisticated technes as needed. But most of the time, for the speed and complexity of the business, a like