I think that he should have spent more time and effort on understanding the question and the availability of data. Why do we need a model? To predict number of merchants? For some forecast? Or we need to find the best possible merchants and try to acquire them? These are very different tasks. Also how will they fit the model, on what data? We don't have any historical information that is tied to the outcome - number of merchants or propensity. Does it mean we'll need to use a combo of models from the similar region and may be some simulation? Once we build the model, How do we evaluate it? We don't have any historical data to validate it. May be some sort of experiment like diff in diff?
This was a huge mess, the intial response to defining sucess metric lacked depth. A fresh grad can come up with # of orders or revenue. He also then rambleed about varius data without framing the exact model and how it would be used to reach a final business decision. Lots of high level fluff without actually answering the underlying DS question.
I think that he should have spent more time and effort on understanding the question and the availability of data. Why do we need a model? To predict number of merchants? For some forecast? Or we need to find the best possible merchants and try to acquire them? These are very different tasks. Also how will they fit the model, on what data? We don't have any historical information that is tied to the outcome - number of merchants or propensity. Does it mean we'll need to use a combo of models from the similar region and may be some simulation? Once we build the model, How do we evaluate it? We don't have any historical data to validate it. May be some sort of experiment like diff in diff?
This was a huge mess, the intial response to defining sucess metric lacked depth. A fresh grad can come up with # of orders or revenue. He also then rambleed about varius data without framing the exact model and how it would be used to reach a final business decision. Lots of high level fluff without actually answering the underlying DS question.
Why did DD saturate the Denver, CO Market with hiring Dashers with in turn this cut the pay of all dasher in the area,,,, Was this you Hanif?????
Why only Softball questions Jay,,,, ---- Let's talk about Dasher obstacles with DD...
Super informative session. Thanks for sharing it.
This is helpful!
Always good to see you Hanif ;)
Thanks Jitesh. Great to hear from you.