[in case you missed it] Data Science Weekly - Issue 257
Issue #257 Oct 25 2018
Editor Picks
A neural network designs Halloween costumes
It’s hard to come up with ideas for Halloween costumes, especially when it seems like all the good ones are taken. And don’t you hate showing up at a party only to discover that there’s *another* pajama cardinalfish? So, I wanted to find out if a neural network could help invent Halloween costumes...
Uber’s Big Data Platform: 100+ Petabytes with Minute Latency
In this article, we dive into Uber’s Hadoop platform journey and discuss what we are building next to expand this rich and complex ecosystem...
How to build your own AlphaZero AI using Python and Keras
Teach a machine to learn Connect4 strategy through self-play and deep learning...
A Message from this week's Sponsor:
Learn data science now, pay after you get a job.
If you’re looking for more advanced learning in data science, check out Springboard’s Data Science Career Track. With one-on-one mentorship and career support, the Data Science Career Track is designed to help you master job-ready skills and actually land your dream job. To make the Career Track more accessible, Springboard now has a new payment plan: deferred tuition.
With Springboard's Deferred Tuition Plan, you can enroll in the Data Science Career Track for a small upfront deposit, with no additional payment until you complete the course and start a new job! If you stick to the plan, but don’t find a job, the course is free.
Apply today!
Data Science Articles & Videos
How Becoming Not a Data Scientist Made Me a Better Data Scientist
Talk by Joel Grus, Research Engineer at Allen Institute for Artificial Intelligence...
Highlights from RecSys 2018
Here is a summary of the recent Conference on Recommender Systems I wrote with my Spotify colleagues Zahra Nazari and Ching-Wei Chen...
Digital immortality:
How your life’s data means a version of you could live forever
Your family and friends will be able to interact with a digital “you” that doles out advice—even when you’re gone...
Trellis Networks for Sequence Modeling
We present trellis networks, a new architecture for sequence modeling. On the one hand, a trellis network is a temporal convolutional network with special structure, characterized by weight tying across depth and direct injection of the input into deep layers. On the other hand, we show that truncated recurrent networks are equivalent to trellis networks with special sparsity structure in their weight matrices. ...
Eta Compute Debuts Spiking Neural Network Chip for Edge AI
Chip can learn on its own and inference at 100-microwatt scale, says company at Arm TechCon...
deeplabv3
PyTorch implementation of DeepLabV3, trained on the Cityscapes dataset...
Partial Convolutions for Image Inpainting using Keras
Keras implementation of "Image Inpainting for Irregular Holes Using Partial Convolutions"... it's been a great learning experience for me to implement the architecture, the partial convolutional layer, and the loss functions...
Curiosity and Procrastination in Reinforcement Learning
Reinforcement learning (RL) is one of the most actively pursued research techniques of machine learning, in which an artificial agent receives a positive reward when it does something right, and negative reward otherwise. This carrot-and-stick approach is simple and universal, and allowed DeepMind to teach the DQN algorithm to play vintage Atari games and AlphaGoZero to play the ancient game of Go. This is also how OpenAI taught its OpenAI-Five algorithm to play the modern video game Dota, and how Google taught robotic arms to grasp new objects. However, despite the successes of RL, there are many challenges to making it an effective technique...
Jobs
Data Scientist - Hearts & Science - NYC
As every CMO knows, employing technology, data and analytics to business challenges is the cost of entry, not a nice-to-have. Unlocking the value of consumer data can deliver growth, revenue and ROI, and create differentiation to stay on top. Real-time action is instrumental in driving results within a highly fragmented marketplace. Hearts & Science’s Marketing Science solutions are built to deliver insightful results with speed, accuracy and a single version of the truth. In leveraging the DNA of a marketing agency with a talent pool of developers, data scientists and Ph.Ds, we hold a unique position in the increasingly crowded ad tech and consulting space...
Training & Resources
Train A One Layer Feed Forward Neural Network in TensorFlow With ReLU Activation
Learn how to train a one layer feed forward neural network in TensorFlow with ReLU activation, Softmax Cross Entropy with Logits, and the Gradient Descent Optimizer, via a screencast video and full tutorial transcript...
Computer Vision for Model Assessment
Talk describes using computer vision to automate statistical model assessment for machine learning...
Trellis Networks for Sequence Modeling
We present trellis networks, a new architecture for sequence modeling. On the one hand, a trellis network is a temporal convolutional network with special structure, characterized by weight tying across depth and direct injection of the input into deep layers. On the other hand, we show that truncated recurrent networks are equivalent to trellis networks with special sparsity structure in their weight matrices. ...
Books
Data Visualization with Python and JavaScript:
Scrape, Clean, Explore & Transform Your Data Learn how to turn raw data into rich, interactive web visualizations with the powerful combination of Python and JavaScript. With this hands-on guide, author Kyran Dale teaches you how build a basic dataviz toolchain with best-of-breed Python and JavaScript libraries—including Scrapy, Matplotlib, Pandas, Flask, and D3—for crafting engaging, browser-based visualizations...
For a detailed list of books covering Data Science, Machine Learning, AI and associated programming languages check out our resources page.
P.S., Want to reach our audience / fellow readers? Consider sponsoring - grab a spot now; first come first served! All the best, Hannah & Sebastian