Felix Wick

[Felix Wick @WickFelix]

Try it out for your probabilistic, ML-based decision making!

[Felix Wick @WickFelix]

In our latest Cyclic Boosting release https://t.co/hsMftLtO2F, we introduce quantile regression (via pinball loss) and subsequent (requiring 3 predicted quantiles) fit-less estimation of full individual probability distributions by means of quantile-parameterized distributions.

[Felix Wick @WickFelix]

brief summary of main ML concepts: https://t.co/ofv5xU2Up9

[PyConDE & PyData @PyConDE]

We just open-sourced Cyclic Boosting, a pure-Python ML algorithm that's explainable, accurate, robust, easy to use, and fast! Learn more in our presentation #Cycl… @wickfelix at #PyConDE #PyDataBerlin
https://t.co/4O8kGZf7Zn

[Felix Wick @WickFelix]

First open-source pre-release (still lots of polishing needed) of the Cyclic Boosting ML algorithms: https://t.co/mAWMRsnNL7
Feel free to try it out, simply do: pip install cyclic-boosting
Please come back with criticism and suggestions. Contributors highly welcome!

[The Cultural Tutor @culturaltutor]

A short history of the world in 13 maps & infographics:

1. Everybody alive today compared to everybody who has ever lived

[Felix Wick @WickFelix]

my recommendation for an off-the-shelf ML algorithm for regression on structured data: https://t.co/McFx71X4iv

[Don Winslow @donwinslow]

This is legit ninja shit.

[Felix Wick @WickFelix]

Evolution provided us with moderate math skills, AI to the rescue. https://t.co/CKUJxBeehX

[Felix Wick @WickFelix]

Just stumbled upon this really nice story from @andrey_kurenkov about the history of neural networks: https://t.co/LTUtSw56Ou

[Felix Wick @WickFelix]

@antgoldbloom For products and locations, you need to find a way to deal with categorical features of high cardinality though. Simple one-hot encoding is not great and especially tree-based methods suffer here.

[Felix Wick @WickFelix]

@antgoldbloom A single model used across all SKUs is not only better in terms of learning commonalities across SKUs, but also operationally way more convenient.

[Felix Wick @WickFelix]

@antgoldbloom One of the most important, but usually overlooked, issues (especially for short-term forecasting) is temporal confounding. Autocorrelation is spurious and can mask true causal effects, e.g., from promotions or events, for the model. Take care how you include it.

[Felix Wick @WickFelix]

@antgoldbloom This is better in terms of variance, but you pay for it with bias. In general, best is to learn on the granularity you want to predict.

[Felix Wick @WickFelix]

@PengM83 @antgoldbloom For new products (and locations) use attributes (including product groups if hierarchy available) and go for embeddings.

[Felix Wick @WickFelix]

@marktenenholtz Probably easiest to have a look at section 5 in https://t.co/66mi4VRZQs or this presentation: https://t.co/tFQyHmOVVH

[Felix Wick @WickFelix]

@marktenenholtz But beware of the detrimental effects of temporal confounding.

[Felix Wick @WickFelix]

@marktenenholtz Two more options: Learn as a distinct category or ignore by setting to a neutral value (if your algorithm allows that).

[Felix Wick @WickFelix]

@peterhoffmann @rasbt @BlueYonder Yes. But in fact, we recently made a twist toward upside-down reinforcement learning.

[Felix Wick @WickFelix]

Was it really just symmetry breaking via setting random initial weights what Rosenblatt missed to get backpropagation to work? (as described in Genius Makers)