search

mattskier / ravens-dse-511

DSE_511 Group Project, Baltimore Ravens
2 stars 2 forks source link

Regression analysis #13

Open mattskier opened 3 years ago

mattskier commented 3 years ago

Regis wrote code, rest should analyze and comment if needed (can talk on slack), trying to predict loan amount

mattskier commented 3 years ago

looks good, think an RMSE of 0.06 may be close to the best we can get without overfitting on the set we have, potentially check if logistic regression works as well?

nisengweregis commented 3 years ago

Thanks, Matt! I can look into logistic regression as well. Since the amount of loan is a continuous variable, I was wondering if logistic regression would be appropriate as it is usually applicable to a binary response variable. What do you think?

Regis, Nisengwe | Read my blog posts

Le vendredi 4 décembre 2020 à 10:41:45 UTC−5, Matt Horan <notifications@github.com> a écrit :

looks good, think an RMSE of 0.06 may be close to the best we can get without overfitting on the set we have, potentially check if logistic regression works as well?

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub, or unsubscribe.

mattskier commented 3 years ago

There’s a potential that loan amounts may increase on an exponential basis though (not much difference between a 100k loan and a 120k loan in risk profile but a 20k is twice as much as an 10k loan), we can see if it goes anywhere

On Dec 4, 2020, at 11:54 AM, Regis, Nisegwe notifications@github.com wrote:

 Thanks, Matt! I can look into logistic regression as well. Since the amount of loan is a continuous variable, I was wondering if logistic regression would be appropriate as it is usually applicable to a binary response variable. What do you think?

Regis, Nisengwe | Read my blog posts

Le vendredi 4 décembre 2020 à 10:41:45 UTC−5, Matt Horan notifications@github.com a écrit :

looks good, think an RMSE of 0.06 may be close to the best we can get without overfitting on the set we have, potentially check if logistic regression works as well?

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub, or unsubscribe. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.