100+ free data science books: https://
bit.ly/3ApVZdJ credit: @marcusborba
@mit_csail
-
100+ Free Data Science Books for Learning
By
–
-
Finch Compiler Optimizes Computations with Looplets Language
By
–
New compiler Finch adapts computations to the structure of data automatically w/language of basic loop building blocks called Looplets. It enables new loop optimizations across multiple domains, unifying techniques such as sparse tensors & databases: https://
bit.ly/3mmoxU3 -
MIT’s Liquid Neural Networks Enable Autonomous Drone Navigation
By
–
Published today in @SciRobotics
, MIT researchers showed that a class of AI systems, dubbed “liquid” neural networks, can understand the task they’re given in autonomous drone navigation experiments. Paper: https://
bit.ly/3Aai6GW @ramin_m_h -
MIT and Financial Leaders Develop Accountable Personal Data Management Systems
By
–
MIT researchers collaborate with Capital One, Fidelity, MassMutual, and Visa on creating accountable, traceable systems for respectful and lawful management of personal financial data. Press release: https://
bit.ly/3LarXmI
White paper: https://
bit.ly/41AZNq5 -
Network filtering algorithm detects hierarchical community structures in data
By
–
New network filtering algorithm can be used to find hierarchical community structure in various types of data. The algorithm can detect structures in financial time series data sets as well as image and electrocardiogram data. Paper: https://
bit.ly/3JkkeQB -
LISP Programming Language Celebrates 66th Birthday Anniversary
By
–
Happy birthday to LISP, first unveiled in a 1960 paper written @MIT
: https://
stanford.io/1FA4PWs (credit: @xkcdComic
) -
How to Confuse Your Machine Learning Model
By
–
How to confuse your machine learning model. (credit: @teenybiscuit
) -
Project Complexity: First Steps to Underestimating Difficulty
By
–
“The first step of any project is to grossly underestimate its complexity and difficulty.” — Nicoll Hunt @nicollhunt
-
MIT Develops BuildIt and D2X Tools for Custom Language Creation
By
–
MIT researchers created “BuildIt”, a tool for non-compiler experts to easily create languages to solve performance intensive problems. “D2X” automatically retrofits existing debuggers for the new languages: https://
bit.ly/41j1Wql
