Quasi-KL Divergence (QKL) Dropout, a stochastic regularisation technique for training of neural networks, has recently been reinterpreted as a specific type of approximate inference algorithm for Bayesian neural networks. The main contribution of the reinterpretation is in providing a theoretical framework useful for analysing and extending the algorithm. We show that the proposed framework suffers from several issues; from undefined or pathological behaviour of the true posterior related to use of improper priors, to an ill-defined variational objective due to singularity of the approximating distribution relative to the true posterior. Our analysis of the improper log uniform prior used in variational Gaussian dropout suggests the pathologies are generally irredeemable, and that the algorithm still works only because the variational formulation annuls some of the pathologies. To address the singularity issue, we proffer Quasi-KL (QKL) divergence, a new approximate inference objective for approximation of high-dimensional distributions. We show that motivations for variational Bernoulli dropout based on discretisation and noise have QKL as a limit. Properties of QKL are studied both theoretically and on a simple practical example which shows that the QKL-optimal approximation of a full rank Gaussian with a degenerate one naturally leads to the Principal Component Analysis solution. …

“Analysts will need a proper understanding of math, statistics, algorithms, and other related sciences in order to deliver meaningful results. They must pair that theoretical knowledge with a firm grasp of the modern-day tools that make the analyses possible. That means having an ability to express queries in terms of MapReduce or some other distributed system, an understanding of how to model data storage across different NoSQL-style systems, and familiarity with libraries that implement common algorithms.” Q Ethan McCallum, Ken Gleason ( 2013 )

Data science is one of the hottest topics in the 21st century because we are generating data at a rate which is much higher than what we can actually process. A lot of business and tech firms are now leveraging key benefits by harnessing the benefits of data science. Due to this, data science right now is really booming.

We talk about geographic bubbles a lot these days. Some areas are isolated, in their own bubble. Other areas seem more connected. Emily Badger and Quoctrung Bui for The Upshot looked at this geographic connectedness through the lens of Facebook friendships.

Language identification (LI) is the problem of determining the natural language that a document or part thereof is written in. Automatic LI has been extensively researched for over fifty years. Today, LI is a key part of many text processing pipelines, as text processing techniques generally assume that the language of the input text is known. Research in this area has recently been especially active. This article provides a brief history of LI research, and an extensive survey of the features and methods used so far in the LI literature. For describing the features and methods we introduce a unified notation. We discuss evaluation methods, applications of LI, as well as off-the-shelf LI systems that do not require training by the end user. Finally, we identify open issues, survey the work to date on each issue, and propose future directions for research in LI. Automatic Language Identification in Texts: A Survey

As an avid podcast listener, I had always wished for an easy way to glimpse at the transcript of an episode to decide whether I should add it to my playlist (not all episode abstracts are equally helpful!). Another challenge with podcasts is that, although they contain a wealth of knowledge that is often not available in blogs and other text formats, there’s isn’t a readily available search engine like Google to index and search the content. What if we could build a tool that converts the audio to text and then build a searchable index on all of our favorite podcast feeds so users could discover information that interests them, without having to listen to a full episode?

Shiny Applications Internationalization (shiny.i18n)It provides easy internationalization of Shiny applications. It can be used as standalone translation package to translate reports, interactive visuali …

This came up in response to a student’s question.

Last weekend we had another fine PyConUK (2018) conference. Each year the conference grows, the Django Girls group had 70 or so women learning Django (and, often, Python for the first time). The kids hack day was a great success. The Pythonic-hardware demo session was fun.

Help! I can’t reproduce a machine learning project!