Do Tran

Sr. Data Scientist @TetraPak, ex @Inria. Tweet about #AI, #MachineLearning, #BigData, #Optimization, #Web, #Cloud, #DevOps. 🇸🇪 🇫🇷

[scikit-learn @ La Fondation Inria @sklearn_inria]

We are hiring our open-source and community manager:
https://t.co/nJeadMG6s7

Come help develop @scikit_learn , it's community, and the foundation's partners.

We want someone into people as much as code. Someone who likes to inspire and cares about open sources

[William Falcon ⚡️ @williamfalcon]

@PyTorchLightnin has become a favorite for AI researchers and industry experts... but we thought there was a high-level framework (lightning is mid-level), missing for common use cases.

Excited to introduce Flash!
Blog:https://t.co/RkQIWADDiL
Repository: https://t.co/soVQXdo4nB

[Xavier Bresson @xbresson]

My main talks on Graph Neural Networks in 2020

1. Introduction to GNNs
https://t.co/TmkJwFF3ds

2. Recent developments in GNNs
https://t.co/mter34Qutg

3. Benchmarking GNNs
https://t.co/fXz2zOllAC

Hope they can be useful.
Happy new year to everyone !

[Andrew Ng @AndrewYNg]

What rules regarding publishing papers would be fair, when it relates to work done by researchers working for companies? I ask this question in this week's The Batch, and would love to hear your thoughts. https://t.co/u97lUGjH6W

[PyTorch @PyTorch]

PyTorch Lightning V0.9 is available now featuring the final API with better data decoupling, shorter logging syntax, synchronized batchnorm, and more. Learn more: https://t.co/8zbEGGCaYJ

[Jeremy Howard @jeremyphoward]

Today is https://t.co/GEOZuodrZj’s biggest day in our four year history. We are releasing our new course, fastai v2, fastcore, fastscript, fastgpu, and the official launch of the book. 1/
https://t.co/odaoVM75o0

[Behnam Neyshabur @bneyshabur]

💡💡What is the best acc an MLP can get on CIFAR10❓

65%❓ No, 85%‼️

Trying to understand convolutions, we look at MDL and come up with a variant of LASSO that when applied to MLPs, it learns local connections and achieves amazing accuracy!

Paper: https://t.co/PUb2Q4tIBT

1/n

[Ed H. Chi @edchi]

I didn't discover this feature until today. This should be the default at every presentation!

https://t.co/rFlafC5pGl

[Kevin Markham @justmarkham]

scikit-learn 0.22 is out! 🎉🎉

- feature importance for *any* model
- gradient boosting models support *missing values*
- KNN-based imputation
- built-in model stacking
- tree pruning
- new plotting API

Highlights: https://t.co/bsrSQOQZTC

#Python #MachineLearning #DataScience

[Jeremy Howard @jeremyphoward]

This is a really well-done article with lots of great examples and clear explanations, showing how to use dropout to get prediction confidence levels at inference time

[Do Tran @dotranet]

“Deep learning has a size problem” by Jameson Toole https://t.co/nlHWkTFLG2

[Kevin Markham @justmarkham]

🐼🤹‍♂️ pandas trick:

Want to explore a new dataset without too much work?

1. Pick one:
➡️ pip install pandas-profiling
➡️ conda install -c conda-forge pandas-profiling

2. import pandas_profiling
3. df.profile_report()
4. 🥳

See example 👇

#Python #DataScience #pandastricks

[François Fleuret @francoisfleuret]

My deep-learning course materials are now updated to PyTorch 1.0.0.

https://t.co/a3AaGEG3Bq

Close to 900 slides! Comments are welcome.

#deeplearning #idiap #epfl #pytorch

[Do Tran @dotranet]

This New Google Technique Help Us Understand How Neural Networks are Thinking https://t.co/2FFBtyBXUP

[Geoffrey Hinton @geoffreyhinton]

This is a much better version of capsules.

[Yann LeCun @ylecun]

ML/CV publication venues are moving up in the h5-index ranking:
#9: CVPR (240) ahead of PNAS, PRL, JAMA.
#27: NeurIPS (169)
#42: ICLR (150) ahead of Neuron and Nature Neuroscience
#56: ECCV (137)
#59: ICML (135)
#71:... https://t.co/WzNDqAIZc8

[Peter Welinder @npew]

Data classes in Python 3.7 are great for encapsulating hyper parameters in machine learning models. Gone are long keyword argument lists. Much clearer code.

[Vincent Le Moign @webalys]

Streamline 3.0 represents 1 year of work between 7 designers — over 10,000 hours of combined work. 🤜💥🤛
With over 30,000 icons, it's the world’s largest icon library.

See what's new 👉 https://t.co/OvJx3yarqw

[Vala Afshar @ValaAfshar]

Distance yourself from people who:

1 speak poorly about others
2 have problems for every solution
3 enjoy complaining
4 always feel entitled
5 blame others for losing
6 take more than give
7 make you feel bad about yourself
8 cheat or lie
9 bully or troll
10 nitpick as a hobby