Every two weeks, we find the most interesting data science links from around the web and collect them in Data Science Briefings, the DataMiningApps newsletter. Subscribe now for free if you want to be the first to get up to speed on interesting resources.
- Google Turning Its Lucrative Web Search Over to AI Machines
For the past few months, a “very large fraction” of the millions of queries a second that people type into Google’s search engine have been interpreted by an artificial intelligence system, nicknamed RankBrain.
- How Tesla is ushering in the age of the learning car
Tesla’s new autopilot system is on the cutting edge of machine learning, connectivity and mapping data.
- What a Deep Neural Network thinks about your #selfie
This amusing post describes how you can train a ConvNet to rate your next selfie.
- Generating Captions: Describing Videos with Neural Networks
Automatically describing the content of an image is a fundamental problem in artificial intelligence. This post shows how neural networks can automatically describe the content of videos.
- Python vs. R vs. COBOL: Which is best for Data Science?
A fun, sarcastic read making fun of similar comparison posts.
- Thought Vectors, Deep Learning & the Future of AI
This DL4J article describes the use of thought vectors and their relationship to the better known concept of word vectors.
- Theoretical Motivations for Deep Learning
This post explores the idea that if we can successfully learn multiple levels of representation then we can generalize well. It hence outlines theoretical motivations towards the use of deep learning based techniques.
- Apache SINGA: a general distributed deep learning platform
The Apache Foundation announces the availability of the first release of Apache SINGA, a general distributed deep learning platform.
- An Exact Algorithm for Finding Minimum Oriented Bounding Boxes [pdf]
We end with two interesting papers, the first presents a new method for computing tight-fitted enclosing bounding boxes for point clouds in three dimensions.
- A Faster Cutting Plane Method and its Implications for Combinatorial and Convex Optimization [pdf]
This paper presents a groundbreaking new general purpose optimization algorithm that promises order-of-magnitude speedups for some problems. Implementing this approach in readily-available software packages might still require some time, however, as the technique is not straightforward to implement in practice.
