Walmart project questions


Hi, I have few doubts in the Walmart project session which are as follows - 

1) Why no analysis of columns was done using diagrams? Is it because you already knew the data well or some other reason?
2) Why don't we use polynomial regression in Walmart project?
3) Why no scaling was performed? Is it because you already knew the data well or some other reason?
4) Why did we use Lasso and not Ridge?
5) Why cross tabulation performed only on store and not on other columns? I mean, how can we be sure which is the column to perform cross-tabulation on it.
6) How to estimate range of alpha? You started with 10 and went till 20. What if I had chosen 1 to 10. Is it just a random guess or some logic is behind it?
7) In lasso graph of the regression, we can see the errors increase with increasing value of alpha. Does this mean more complex model is introducing more errors? 

Kindly provide answers to them.

1 Answer(s)


Refer to session-5 where we have discussed all of the these questions in details...

And for every project we do necessary analysis .If visualization is required then we go for , In walmart more about analysis rather visualization..


To understand when ,why, how do we go for this methods.. enroll in data science in python or setup a meeting by contacting with Dezyre.

Hope it answer your question