Company Market Cap Prediction (Model Selection) - Data Every Day

[@jiolfabio]

Hello, very good video. But using the mean to fill NaN on the entire dataset does not follow best practices to prevent data leakage. P.S. The market cap is in millions so the min is 43 millions, and the error is huge.

[@kaischmid9118]

But how is the error 40B and the r^2 score is 0.6

[@abc-co7fy]

as always, great Gab ! ♥☺

[@dogus8662]

I was waiting for this moment. when will you use lstm?

[@gcdatkin]

I have covered them in the past! :) Here are some: kzbin.info/www/bejne/e4rClYCOjdmUsMk kzbin.info/www/bejne/d2PEoGatrth0fM0 kzbin.info/www/bejne/fWaud5tvir-BnLs

[@dogus8662]

@@gcdatkin thanks for the return

[@satyamtripathi1732]

Bro u r awsm please its a request make a video on like if u r geeting accuracy in negative how to increase that like u have used many models and ramdomforest giving best accuracy so using hyperparameters increase accuracy please 🙏🏻🙏🏻🙏🏻🙏🏻🙏🏻

[@gcdatkin]

Glad you are enjoying! Be careful not to confuse accuracy with R^2 score. The accuracy metric is used in classification tasks, but in regression tasks we use R^2 score, which measures how much better our model is than just predicting the mean of y_test every time. You can't get a negative accuracy, but you can get a negative R^2. For hyperparameter optimization, check out these ones: kzbin.info/www/bejne/d3mpaKCYd7SIna8 kzbin.info/www/bejne/gpm9d5Sop92VY6c

[@satyamtripathi1732]

@@gcdatkin thnkyou very much i am very happy that u replied

[@ayencoscolfield3312]

But though dont think i understand the phrase "company market cap" in this context and that seems to be the target variable but will follow along, thanks always

[@gcdatkin]

Market cap (short for market capitalization) is the total market value of all outstanding shares for a publicly traded company.

[@ayencoscolfield3312]

@@gcdatkin ok thanks