Vincent Warmerdam: Winning with Simple, even Linear, Models | PyData London 2018

[@SUGATORAY]

“If you understand your solution better than the problem, then you are doing something wrong.” 👏👏

[@gautame]

But that sounds like an argument against "simple" models.

[@christianlagares3865]

⁠I wouldn’t say it’s an argument against simple models. Rather, if you understand the fundamental aspects of your problem then the solution will often avoid unnecessary complexity.

[@jaydeepradadia963]

00:00 Intro 01:29 Topics to be covered 02:02 XOR problem 04:45 Time-series trick 10:20 Weighted Linear Regression 17:23 Passive-Agressive Algorithms 21:42 Recommender Systems 26:20 Example of Video Games 31:33 Example of Chickens 35:26 Conclusion

[@yashgupta471]

Someone should pin this.

[@MMarcuzzo]

You might caption the video, I guess.

[@marco_gorelli]

The RBF features tip at 6:10 is really useful for modelling holidays, where the effect isn't exactly binary - thanks Vincent!

[@juliocezarsilva5979]

This definitely changed my life

[@TraininData]

This talk makes me fall in love with feature engineering all over again! Today more relevant than ever, with new regulation coming in, keeping our models simple and interpretable is a must.

[@dennisestenson7820]

This is great info! Too bad it was so rushed. I'd have liked to see more in depth examples.

[@matattz]

Hey, do you have any book recommendations to learn exactly this stuff he is talking about? This is really interesting

[@dangernoodle2868]

Not the author, but I may recommend reading "Causal Inference for the Brave and True", with the radial basis functions he was doing a kind of "synthetic control" and maybe finding a resource on splines. For the later portions on machiene learning you can tell he's read "Statistical Rethinking".

[@matattz]

@@dangernoodle2868 looks promising, thank you!

[@igormichetti]

one of the best ai presentation out there

[@DuBoisEdmund-r1t]

White David Garcia Angela White Steven

[@jaydeepradadia963]

By applying a linear mapping to a non linear basis function, non-linearity can be modeled.

[@PotentialEn3rgy]

10/10 thank you!

[@umitkaanusta]

One of the best I've seen

[@jbs3144]

This is a great presentation. Keep it simple and also the suits need to understand.

[@dwipaal-farisi4107]

Love it

[@n.w.4940]

I already like this vid at minute 2

[@ewg6200]

Are you on speed? Slow down. Breathe.

[@gautame]

This talk is a bit dated. Just because it's easy to understand out-of.the-box, doesn't mean that it is better. I'd rather peer into or interrogate a more accurate deep-learning model to understand how it's working than be satisfied by easy-to-generate plots of a simple and less accurate model.