In my post on Recurrent Neural Networks in Tensorflow, I observed that Tensorflow’s approach to truncated backpropagation (feeding in truncated subsequences of length n) is qualitatively different than “backpropagating errors a maximum of n steps”. In this post, I explore the differences, implement a truncated backpropagation algorithm in Tensorflow that maintains the distribution between backpropagated errors, and ask whether one approach is better than the other.

Suppose you have a cluster. Suppose you would like to monitor your cluster as soon as possible without installing all kind of tools on the cluster. A new software package named ISA has been created which can do centralized monitoring for you! This article is a walkthrough for ISA and helps you setting up monitoring for your cluster in just a few minutes.

Scala is jet another programming language in the world of programming languages. Its first version was released in 2001 and was conceptually developed by Martin Odersky, a professor at the EPFL in Switzerland. Fun fact: one of his first projects was called Pizza and was a super set of the Java language. Later on, he developed Scala. The name and logo are easy to explain. First of all, Scala stands for SCalable LAnguage. Another translation: “Scala” means “staircase” in Italian. The logo is based on a particular staircase in one of the buildings of EPFL. Now we have had a decent introduction to Scala, we can start exploring the language itself. So what is Scala all about?

Organizations aspiring to become data-driven need to take a close look at their HR practices. If your company’s hiring and retention standards aren’t keeping up with the times, you may be losing valuable job candidates and employees. To minimize the pitfalls of building a data-savvy workforce, consider these tips.

The following post (Part 2 of two parts) by Vik Paruchuri, founder of data science learning platform Dataquest, offers some detailed and instructive insight about data science workflow (regardless of the tech stack involved, but in this case, using Python). We re-publish it here for your convenience.

This year my team at Yandex organized MLHEP (Machine Learning in High Energy Physics) summer school in Lund, Sweden.

Whenever possible, substitute constructions out of known entities for inferences to unknown entities Occam’s razor (Russell’s version)

This is the first in a series of posts about recurrent neural networks in Tensorflow. In this post, we will build a vanilla recurrent neural network (RNN) from the ground up in Tensorflow, and then translate the model into Tensorflow’s RNN API.

Hackathons are not alike