Skip to content

DSB #125


let’s roll, it’s Friday and DSB! My personal favorite in this volume is an article about optimization of a pokemon team from Analytical. Of course, the article itself is quite simple, but never mind that.

As always, enjoy your reading.

Analytical – How to optimize a pokemon team? Author is using the PuLp library for linear programming and optimization. – How to read and interpret SHAP. – Comprehensive guide to modeling proportions with Bayesian beta and zero-inflated beta regression models.

Computer Science & Science – PostgreSQL 14 was released a few weeks ago, so learn about known features of PostgreSQL. And there are many of them. – EDA (event driven architecture) is here for your API to handle complex asynchronous solutions. (rcmd by reader) – Personally, I hate GUI. But you should at least know Bamboolib. Because it will make from your jupyter notebook a GUI for Pandas. Uff. (rcmd by reader)

Graphs and Visualizations – Everyone knows about UN Climate Change Conference in Glasgow, so have a look who is the “best country” in the world in CO2 emission per capita. And yeah, we are pretty good… (rcmd by reader) – Scientific Python — Volume II: Scientific Visualization – Python & Matplotlib is an free-to-read book that will teach you how to how to tweak your boring matplotlib charts. Coming soon! (rcmd by reader) – Funny simulation of zombie outbrake on a map. (rcmd by reader)

Business and Career – A view of one quant on Zillow, a company that is an example of how one prediction model might almost ruin the whole company. – How to measure analytical work? And why is it so difficult? – Everybody
trusts in DeFi nowdays. How will decentralized finance affect banking or broadly the whole society?

Pop DSB #120 bylo úžasné instruktážní video z roku 1953, jak nastavit námořní dělo. Samotná technika zaměřování v sobě snoubí dřevní počátky computer science a modelování, takže stojí za to tomu věnovat pozornost. Více se dozvíte v článku. (rcmd by reader) – DERL is a simulation of Darwinian evolution and it shows some promissing results – like existance of Baldwin effect. (rcmd by reader) is a new AI-driven search engine. Should Google be afraid?

Education – Tohle je hodně dobrý. Jednoduchý průvodce AI pro širokou veřejnost na mnoho rozličných témat. Neuvěřitelný z jakých zdrojů můžou dneska nejen děti a studenti čerpat. (rcmd by reader) – Outlier detection vs drift detection. Nice and easy to understand article. – Older introduction to computer vision by University of Washington from 2018.

Data & Libraries – API for building distributed applications. (rcmd by reader) – SCENIC is a Python library written in JAX for computer vision created by Google. Should be easy to use.
GitHub is here and paper is here. (rcmd by reader) – Tutorial on ipywidgets in Jupyter Notebooks. And it’s a really good feature. (rcmd by reader)

MLOps – Interesting discussion about cloud/docker workflow. (rcmd by reader) – Why is Kuberenetes sometimes very painful and why it’s not suitable for batch jobs. – How to improve a ML system. This is the first part and is about broken abstraction.

Video & Podcast –  Výborný díl Datacastu s Karlem Šimánkem z o budování AI řešení v cloudu. (rcmd by reader) Minute Papers is a channel that is self explanatory. This video is about AI learning fighting moves. (rcmd by reader) – Deep Learning with PyTorch in 5 hours long video 😀

Papers & Books – How is it with 30 million missing girls in China because of China’s one-child policy? Well, they are probably not missing at all, sometimes data are just wrong. (rcmd by reader) – On November 23rd will be released the second edition of Deep Learning with Python. (rcmd by reader) – Is it better to have more parameters or more labeled examples? It depends on the task. More in paper. (rcmd by reader)

Behind the Fence – Analyst II, Data Science in Liberty Mutual Insurance, Somerville, MA, USA.

Joke (rcmd by reader)

Be First to Comment

Leave a Reply