[in case you missed it] Data Science Weekly - Issue 280
Issue #280 Apr 4 2019
Editor Picks
DeepMind and Google: the battle to control artificial intelligence
Demis Hassabis founded a company to build the world’s most powerful AI. Then Google bought him out. Hal Hodson asks who is in charge...
Putting virtual glasses on a real face: Warby Parker’s Virtual Try-On algorithm
By simulating the real-life process of placing a pair of frames on the face — and taking into account how unique facial features interact with glasses — the tool can figure out just where they will sit. In this post, I’ll give you a glimpse into how we, the Warby Parker team, made this happen...
The Animal-AI Olympics is going to treat AI like a lab rat
The $10,000 competition will test AI with challenges that were originally designed to test animal cognition—to see how close we are to machines that have common sense...
A Message from this week's Sponsor:
Find A Data Science Job Through Vettery
Vettery specializes in tech roles and is completely free for job seekers. Interested? Submit your profile, and if accepted onto the platform, you can receive interview requests directly from top companies growing their data science teams.
Get started.
Data Science Articles & Videos
MIT is using AI to invent new flavor combinations and foods –
and it suggested a shrimp, jelly, and sausage pizza
McCormick will be using 40 years of data from IBM Research to learn about and produce new flavor combinations, with a view to becoming well-versed in flavor palettes, consumer preferences, and sensory science...
Alexa AI scientists reduce speech recognition errors up to 22% with semi-supervised learning
Amazon’s Alexa Speech group scientists today announced they have used what they believe to be one of the largest unlabeled data sets ever assembled to train an acoustic model and improve the intelligent assistant’s ability to understand the human voice...
Personalized Recommendations for Experiences Using Deep Learning
As the world’s largest travel site, TripAdvisor provides a platform for billions of users to research, book, and review their trips across the world. In this blog post, we will explain how our newly-developed ‘Recommended For You’ (RFY) model generates personalized recommendations on our website using users’ browsing history and deep learning. The model has already been tested in production and demonstrated lifts in user engagement and bookings...
The grim reality of life under Gangs Matrix, London's controversial predictive policing tool
AI and machine learning software was meant to make policing fairer and more accountable – but it hasn't worked out that way...
The Intuition behind Adversarial Attacks on Neural Networks
Are the machine learning models we use intrinsically flawed?...
Combining online and offline tests to improve News Feed ranking
A/B testing is an important part of the product improvement cycle for machine learning (ML) technologies. But applying advanced techniques such as Bayesian optimization to optimize these systems can be challenging due to resource limitations. We propose a new multitask Bayesian optimization approach that combines observations from online A/B tests with a simple offline simulator. This model allows for jointly optimizing up to 20 parameters with as few as 40 total online A/B tests...
Neural Machine Translation With Attention Mechanism
Hello guys, spring has come and I guess you’re all feeling good. Today, let’s join me in the journey of creating a neural machine translation model with attention mechanism by using the hottest-on-the-news Tensorflow 2.0...
Diagnosing and Enhancing VAE Models
Although variational autoencoders (VAEs) represent a widely influential deep generative model, many aspects of the underlying energy function remain poorly understood. In particular, it is commonly believed that Gaussian encoder/decoder assumptions reduce the effectiveness of VAEs in generating realistic samples. In this regard, we rigorously analyze the VAE objective, differentiating situations where this belief is and is not actually true. We then leverage the corresponding insights to develop a simple VAE enhancement that requires no additional hyperparameters or sensitive tuning...
White Paper
Preparing your own ML training data?
It takes a lot of data to train an algorithm that is production-ready. For an algorithm to be highly accurate, you absolutely need ata that is labeled and annotated with nearly flawless quality. If you’re not sure what it takes to produce enormous volumes of quality training data, you need the Blueprint....
Want to post here? Email us for details >> team@datascienceweekly.org
Jobs
Data Scientist (Analytics) - Pear Therapeutics - San Francisco or Boston
At Pear Therapeutics, we have the privilege of building the world’s first-ever class of prescription digital therapeutics. By nature of our therapeutics as digital applications, we have access to rich datasets and unique opportunities to drive clinical outcomes. As a Data Scientist, you will be responsible for shaping and delivering data-driven insights. We are looking for data scientists with a deep product sense, who have an innate curiosity, and are eager to dive into large, complex datasets and create actionable insights...
Want to post a job here? Email us for details >> team@datascienceweekly.org
Training & Resources
Infer Dimensions While Reshaping A PyTorch Tensor
Learn how to infer dimensions while reshaping a PyTorch tensor by using the PyTorch view operation, via a screencast video and full tutorial transcript...
Mixture of Variational Autoencoders — a Fusion Between MoE and VAE
In this post I explain how you can use mixture of experts (MoE) and variational autoencoders (VAE) to generate images of digits. Doing so, you can condition on the digit you want to generate an image for. And the beautiful part - no labels are needed :)...
Framework for Understanding Unintended Consequences of Machine Learning
In this paper, we provide a framework that partitions sources of downstream harm in machine learning into five distinct categories spanning the data generation and machine learning pipeline...
Books
Reproducible Research with R and R Studio "a very practical book that teaches good practice in organizing reproducible data analysis and comes with a series of examples..."
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