Ridge Regression | Tikhonov Regularization | Machine Learning #10

[@stefanzottmann1343]

You made it easy to you so much for the professional and simplicity you put in the videos are very underrated.

[@gollum95530]

I've watched dozens of videos on regularization and your explanation is perfect! thanks!

[@grajov123]

One of the best lectures on ridge regression

[@staciarocco8305]

Managed to make this work thanks to your steps Ahmad !!

[@silvacervantez3737]

A machine learning guru !!

[@franco9646]

Keep up the good work, blessing from USA my dear friend :)

[@kaitlynmorgan8441]

Thank you so much for this video, I love you! I hope it helps for the exam tomorrow

[@cucciolatina]

Lovely tutorial ahmad !

[@maryameek5661]

Excellent channel, thank you a lot for your helpful, and well organized videos.

[@harrietkinzer7987]

Awesome lecture and very clear you a lot.

[@dataexpress4399]

Amazing Explanation

[@turanyapraklioglu7759]

Great video this series coming in simple and useful.

[@wendellhaviland5025]

Perfect lesson

[@jamesluna9859]

can't thank you enough :)

[@kylepatrick3526]

what an up-to-point tutorial 🙏🏻

[@jazmynealicia2506]

Thank you Dr

[@MrKean1]

Thank you important lecture.

[@awildacartier1418]

I wish you worked as a professor in where I study.

[@francinasylvie9447]

Thank you for this video, it's so helpful! I can't believe, it's only 1200 consider Patreon account that people could reward you and thank you for your work.

[@Mel32G]

Please make a video on elastic net regressors.

[@MyCocomartin]

The formula appearing in you please do a lecture to prove it ? I am having a hard time deriving you so much Ahmad.

[@dataexpress4399]

nice

[@alikucukel3760]

👏👏👏

[@alexfreeman1036]

13:51 does that mean as alpha increases, then so does under fitting ?

[@heathdotty3022]

12:52 what If I set polynomial to false what will be the expected outcome ?

[@latashiatrammel1980]

12:12 I wonder if pipelines will actually speed up perfectly well when training neural networks, but I wonder if this is

[@gladiscerezo3447]

Lectures are coming in smart, have one concern seems like you are a bit tired while you feel okay ?

[@michaelford9569]

I just did a Google search and turns out that Ahmad Bazzi is a signal processing I ask from where do you buy time to publish all those high quality you own a company editing and helping you record, I don’t

[@braxtonisabel1421]

Oh this was recorded in Lebanon ? Aren’t you in France ?

[@yunyizhao8254]

May I ask why you use half of alpha kzbin.info/www/bejne/hqWlpqR3mcejnLs rather than just use alpha?

[@AhmadBazzi]

Well since when computing the gradient (with respect to theta_i), you will end up with the gradient of the MSE term plus that of the penalized term. The gradient of the penalized term would then be alpha x theta_i (instead of 2 x alpha x theta_i) .. This would just make interpreting the gradient much simpler. In theory, you could multiply alpha with whatever you want, you would only be giving more/less importance to the penalty.

[@valerieesther7056]

1:46 “then instead of just minimizing this guy, you would want to minimize the MSE plus a penalty”.This guy literally talks mathematics.

[@michaelr113]

MICHAELR113 IS BETTER

[@michaelr113]

MICHAELR113 IS BETTER

[@michaelr113]

MICHAELR113 IS BETTER

[@michaelr113]

MICHAELR113 IS BETTER

[@michaelr113]

MICHAELR113 IS BETTER

[@michaelr113]

MICHAELR113 IS BETTER

[@michaelr113]

MICHAELR113 IS BETTER