Original article can be found here (source): Deep Learning on Medium
D4S Sunday Briefing #43
A weekly newsletter with the latest developments in Data Science and Machine Learning and Artificial Intelligence.
Mar 22, 2020
Welcome to the 43rd issue of the Sunday Briefing. This week we take a small break from pandemic coverage to focus a bit more on more traditional data science and machine learning, the notable exception being a wonderful visual history of pandemics by the World Economic Forum that we can put under the heading of visualization and infographics
This week, Google Research announced the release of AutoML-Zero, a new framework to Evolve Machine Learning Algorithms. On the neural network front, we look at Survival Analysis in Keras and data extraction from financial documents.
From more traditional academia, we have a Survey on Contextual Embeddings and one on Adversarial Learning on Graphs and a deep dive into human mobility by looking at interurban movements and Flow descriptors of human mobility networks.
Finally, in our video of the week Derek Kane gives us a detailed tutorial on Hidden Markov Models, a simple example of a dynamic Bayesian network.
Data shows that the best way for a newsletter to grow is by word of mouth, so if you think one of your friends or colleagues would enjoy this newsletter, just go ahead and forward this email to them and help us spread the word!
The D4S team
Our latest post covers the first part of section 1.3 Probability Theory and Statistics, an overview of some of the fundamental theoretical requirements for the journey ahead. The code for each blog post in this series is hosted by a dedicated GitHub repository for this project: github.com/DataForScience/Causality
Tutorials and blog posts that came across our desk this week.
- A visual history of pandemics [weforum.org]
- Survival Analysis with Deep Learning in Keras [towardsdatascience.com]
- How to use deep learning for data extraction from financial documents [nanonets.com]
- Liquidity modeling in real estate using survival analysis [opendoor.com]
- Google Research: AutoML-Zero [github.com/google-research]
- Neural Networks: Everything you Wanted to Know [towardsdatascience.com]
- How to understand your complex machine learning algorithm, and why you should use SHAP [medium.com/vantageai]
Fresh off the press:
Some of the most interesting academic papers published recently.
- A Survey on Contextual Embeddings (Q. Liu, M. J. Kusner, P. Blunsom)
- A Survey of Adversarial Learning on Graphs (L. Chen, J. Li, J. Peng, T. Xie, Z. Cao, K. Xu, X. He, Z. Zheng)
- Stock price prediction using principal components (M. Ghorbani, E. K. P. Chong)
- AutoML-Zero: Evolving Machine Learning Algorithms From Scratch (E. Real, C. Liang, D. R. So, Q. V. Le)
- Tracking COVID-19 using online search (V. Lampos, S. Moura, E. Yom-Tov, I. J. Cox, R. McKendry, M. Edelstein)
- The effect of interurban movements on the spatial distribution of population in China (J. Ye, Q. Hu, P. Ji, M. Barthelemy)
- Flow descriptors of human mobility networks (D. Pastor-Escuredo, E. Frias-Martinez)
Video of the week:
Interesting discussions, ideas or tutorials that came across our desk.
Hidden Markov Models
Opportunities to learn from us
- Mar 27, 2020 — Deep Learning for Everyone [Register]
- Apr 8, 2020 — Time Series for Everyone [Register] 🆕
- Apr 29, 2020 — Applied Probability Theory for Everyone [Register] 🆕
Thank you for subscribing to our weekly newsletter with a quick overview of the world of Data Science and Machine Learning. Please share with your contacts to help us grow!
Publishes on Sunday.
Copyright © 2020 Data For Science, Inc, All rights reserved.