Introductory reading for people who are new to Sig-ML
"All models are wrong. Some are useful" -George Box
Github is an online collaboration platform based on the Git version control system (1). It let's us share our code with the rest of the world painlessly. Moreover, it lets members of Sigml work on common interests even after they have moved on from the institute.
All programs that we develop are stored here as different repositories. Some repositories grow big enough to have their own team and you can ask to be included in them.
There are two general teams. Born early and Born late. This is essentially second years and first years. Besides that different teams crop up as per repositories and projects.
- Tools
- git (github)
- python
- pandas
- sklearn
- matplotlib
- R
- (I have no clue, someone fill this with a Pull Request?)
- Concepts / skills
- general programming
- simple statistics
- linear algebra
- matrix math
- Resources
- wikipedia
- https://news.ycombinator.com
- stackoverflow
- IRC
Most of us usually hang out in #python and ##machinelearning
- https://github.com/scikit-learn/scikit-learn
- https://github.com/pydata/pandas
- https://github.com/numpy/numpy
- https://github.com/matplotlib/matplotlib
- https://github.com/mwaskom/seaborn
- https://github.com/theSage21/reimagined-chainsaw/blob/master/scripts/mlpr
- https://github.com/dask/dask
- https://github.com/ipython/ipython
- http://jupyter.readthedocs.io/en/latest/
- https://github.com/vlfeat/vlfeat
- https://github.com/dmlc/mxnet/