Bayesian methods for hackers book in a #JupyterNotebook format and an adequate dose of 'classical bayesian way' skepticism is a great read :
https://t.co/4CuqxUZnUs . highly recommended
@nkoder@warszawa ja też mam "poczucie", bo realnego wpływu to chyba nie. Skuteczność rekacji na złe parkowanie w ostatnim czasie: 1/5 inne kategorie: 0/2. Niestety brak SLA.
We will be talking about a deep-learning case-study from https://t.co/UhECJ6iEh1 next Tuesday (17.04) in Warsaw: https://t.co/sA5FDpeLUX - a real-life example, see you there!
Anyone wants to work for https://t.co/UhECJ6iEh1 solving company's Machine Learning challenges? It's pretty damn interesting: https://t.co/JVkIO1gJAX or https://t.co/Tb7T2ntMZZ
Long time since I played with Particle and their great IOT chips. New mesh devices https://t.co/CgOSmfZsml look like a great moment for a refresh. Can't wait July...
“We’re applying brittle optimization techniques to loss surfaces we don’t understand.” - #NIPS2017 test of time awardee Ali Rahimi argues that in ML rigor has decreased even as reproducibility has increased
We’re excited to announce Kaggle’s Data Science for Good program! We're hoping this enables our community to come together and make contributions to tough social good problems with datasets outside of our competitions. Read more here: https://t.co/SUSXf1FaRk