Data Science Weekly - Issue 208
Issue #208 Nov 16 2017
Editor Picks
Software 2.0
I sometimes see people refer to neural networks as just “another tool in your machine learning toolbox”. They have some pros and cons, they work here or there, and sometimes you can use them to win Kaggle competitions. Unfortunately, this interpretation completely misses the forest for the trees. Neural networks are not just another classifier, they represent the beginning of a fundamental shift in how we write software. They are Software 2.0...
A Visual Guide to Evolution Strategies
In this post I explain how evolution strategies (ES) work with the aid of a few visual examples. I try to keep the equations light, and I provide links to original articles if the reader wishes to understand more details...
I built a LinearRegression that can play Pong with me
A naive way of creating something towards AI. It’s pretty cool. Do you remember that vintage video game pong? Just two pads and a ball that wasn’t actually round? It looked a bit like this...
A Message from this week's Sponsor:
Attend EGG2017 The Non-Conforming Data Science Conference
Join us on November 30th in Manhattan for new kind of data science conference. Get real-world examples of and insights on developing data-driven organizations from industry-leading players (like Red Hat, Daimler, Slack, Goldman Sachs, and more). Use discount code 15off for a 15% discount!
Data Science Articles & Videos
10 Minutes of Imaginary Japanese Anime Face
10 Minutes of Imaginary Japanese Anime Face, dreamed by (an 256x256 version) of our #MakeGirlsMoe GAN model that is accepted in #NIPS2017 Workshop for Machine Learning for Creativity and Design!...
This AI Chef Wants to Put You on an Environmentally Conscious Diet
Your next veggie burger might be cooked with some chickpeas, black beans—and maybe a pinch of artificial intelligence...
Classification of Meows and Woofs
I built a model to classify a set of .wav files into sounds made by either a cat or a dog...
CheXNet: Radiologist-Level Pneumonia Detection with Deep Learning
We develop an algorithm that can detect pneumonia from chest X-rays at a level exceeding practicing radiologists...
Deep Learning is Eating Software
When I had a drink with Andrej Karpathy a couple of weeks ago, we got to talking about where we thought machine learning was going over the next few years. Andrej threw out the phrase “Software 2.0”, and I was instantly jealous because it captured the process I see happening every day across hundreds of projects. I held my tongue until he got his blog post out there, but now I want to expand my thoughts on this too...
Exploring Line Lengths in Python Packages
Data-driven ruminations on PEP8's 79 character limit & the mark it leaves on Python packages...
Gaussian Distributions are Soap Bubbles
This post is just a quick note on some of the pitfalls we encounter when dealing with high-dimensional problems, even when working with something as simple as a Gaussian distribution...
Understanding Hinton’s Capsule Networks. Part I: Intuition.
Last week, Geoffrey Hinton and his team published two papers that introduced a completely new type of neural network based on so-called capsules. In addition to that, the team published an algorithm, called dynamic routing between capsules, that allows to train such a network. In this post, I will explain why this new architecture is so important, as well as intuition behind it. In the following posts I will dive into technical details...
Jobs
Machine Learning / AI Architect – Research & Development -
Citrix - Patras, Greece Citrix is expanding its Advanced Analytics team, with seasoned professionals in the ML/AI/Data Science and Security domains. You will join a crack team, with years of history delivering high-quality Analytics products, and a global outreach. You will be collaborating with fellow engineering teams across the globe, on the cutting edge Citrix Analytics Service
Key Responsibilities:Research & develop Machine Learning models for security problems, in the areas of Networking, Application & Data.
Suggest, collect and synthesize requirements and create effective features.
Apply research methodologies to identify the Machine Learning models for the problem at hand
Training & Resources
Save The State Of A TensorFlow Model With Checkpointing
Save the state of a TensorFlow Model with Checkpointing using the TensorFlow saver variable to save the session into TensorFlow ckpt files...
On-Device Conversational Modeling with TensorFlow Lite
Today, we announce TensorFlow Lite, TensorFlow’s lightweight solution for mobile and embedded devices. This framework is optimized for low-latency inference of machine learning models, with a focus on small memory footprint and fast performance...
Hosting an R Shiny Application on Amazon EC2
In the second part of this three part series I will discuss how I host wespasplaypredictor.com using a Shiny app hosted on an Amazon EC2 instance, also using Amazon’s Route 53 DNS service to setup a custom domain name...
Books
Our Final Invention: Artificial Intelligence and the End of the Human Era "Well written exploration of the precipice that we are approaching..."
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. We just have a few slots left in 2017 - grab a spot now; first come first served! Email us for more details - All the best, Hannah & Sebastian