Viewing profile — polkapolka

Did you run into latency problems doing arbitrage this way? Or do you rely on API's?

You bypass the API completely and fill orders "manually".

You are thinking too US-centric here. There are jurisdictions where this is allowed. Also don't stare yourself blind on the numbers. Do you unfairly deny 5 minorities or erroneousl…

The identifying characteristics are signal for the minority, but noisy for genpop. Deployment/engineering constraints call for a single model. Realistic scenario.

Eat your own dogfood. Let them douse their babies in asbestos, drink fracking water, or force them to watch Scientology advertisements.

The ads are a red herring (and uncomfortably visible for both parties). The real activity was on Facebook groups, viralizing anti-immigration and far-left news, controling the narr…

Statistical learning with connectionist architectures is driving current AI at scale. To me, this paradigm is also the most promising: learn from data bottom-up, not from experts t…

Faster and more complex computers does not make you faster at manual programming. Computer vision had this before the DL boom: engineers painfully crafting feature extractors. It w…

Symbolic AI is machine programming. Connectionist AI is machine learning. Machine programming simply does not scale. It is also not biologically plausible: it is not as if God put …

Agreed. Logistic regression with poly kernel or good engineering interactions can equal or beat more complex models for a fraction of the budget. All the more power to you if a sol…

Text data traditionally seen as unstructured. Try a simple MLP.

For an accessible recent overview see: https://christophm.github.io/interpretable-ml-book/

Yeah, it is a good first benchmark. But view interpretability as separate from accuracy. You can explain black box algorithms just fine these days. Logistic regression is high bias…

Sure there are constraint and pros and cons, but still: pick a neural network for unstructured data. Can always unsupervised pretrain and fine tune on a tiny dataset.

The image for logistic regression is hilariously wrong. It shows the sigmoid as a decision boundary. Also dont get hung up about no free lunch theorem. That is a great result in co…

I give Musk a little bit more credit than brutely calling someone a pedo without any proof. If you are as rich as Musk, there must be someone you can pay to dig up dirt on your opp…

Csv and a permitive license like CC0. Provide good meta data to make provenance easier. Mark up your dataset with schema.org.

Gone to sleep for the final time. Ash to ... And dust to dust.

Dennett aims to solve this using heterophenemenology: https://en.m.wikipedia.org/wiki/Heterophenomenology In this framework, utterances can be studied without taking their truth va…

If you spend one year on applying deep learning, you can train a net on a 100 different data sets. That's where the intuition comes from. You'll debug a lot. People with zero exper…

Articles like these should come with a disclosure: "the authors own bitcoin and augur" or something to that effect.