Issue 74: CognitionX Data Science, AI and Machine Learning Briefing

Last night we welcomed Dr. Anil Anthony Bharath (Reader in Image Analysis at Imperial and Co-Founder, Cortexica Vision) to our meet up. He took us on a deep learning journey from 1978 to today discussing XNOR Net, Adversarial networks, N-headed beasts, LTSMs and Neural Turing machines and Jennifer Aniston cell.If you’re excited by this do come along to the next Deep Learning talk on the 3rd of November. If you’re not only excited but understand all of this, get in touch as we have companies in the community looking for deep learning engineers and researchers
Best
Tabitha UntilTheBotsTakeOver Goldstaub

Ethics Questions of the Day

“AI-Ready or Not: Artificial Intelligence Here We Come!”Weber Shandwick report shows those surveyed were six times more likely to see AI’s impact on society as positive than negative and seven times more likely to expect an positive impact than negative on their personal lives. EVEN though the vast majority of consumers expect jobs to be lost (82%) rather than gained (18%) due to AI. Also interesting to note women more positive about the use of AI than men. Check out the full report. Learn more >

Research 

Deep Identity-aware Transfer of Facial AttributesThis paper presents a Deep convolutional network model for Identity-Aware Transfer (DIAT) of facial attributes. Mu Li, Wangmeng Zuo, David Zhang’s DIAT and DIAT-A models can provide a unified solution for several representative facial attribute transfer tasks such as expression transfer, accessory removal, age progression, and gender transfer. I’ll be watching closely for the commercial uses of something like this in the beauty industry. Learn more >

Robots taking humans jobs

When we were children, the rule was always the person who cut the cake had to pick the slice they wanted last.

There is now an algorithm that can fairly divide a cake among any number of people. Taking fair to another level!

Aziz and Mackenzie’s new algorithm builds on an elegant procedure that mathematicians John Selfridge and John Conway independently came up with around 1960 for dividing a cake among three people. Learn more >

Education Advice and Training we rate

Deep Learning Part 1: Comparison of Symbolic Deep Learning Frameworks
This blog series is based on Anusua Trivedi, Sr. Data Scientist – Microsoft’s talks on re-usability of Deep Learning Models at the Hadoop+Strata World Conference in Singapore. Learn more> 

Tools of the trade

Is Artificial Intelligence Permanently Inscrutable?At a DARPA conference early this year, program manager David Gunning summarised the trade-off in a chart that shows deep networks as the least understandable of modern methods. At the other end of the spectrum are decision trees, rule-based systems that tend to prize explanation over efficacy.

In this Nautil article Aaron M. Bornstein (Princeton Neuroscience Institute) explains his view that despite new biology-like tools, some insist interpretation is impossible.  Learn more> 

Open Source

Image Synthesis from Yahoo’s open_nsfwWarning: This post contains abstract depictions of nudity and may be unsuitable for the workplaceYahoo’s recently open sourced neural network, open_nsfw, is a fine tuned Residual Network which scores images on a scale of to on its suitability for use in the workplace. In the documentation, Yahoo notes “Defining NSFW material is subjective and the task of identifying these images is non-trivial. Moreover, what may be objectionable in one context can be suitable in another.” What makes an image NSFW, according to Yahoo? Like Google’s Deep Dream, this visualization trick works by maximally activating certain neurons of the classifier. Unlike deep dream, they optimise these activations by performing descent on a parameterization of the manifold of natural images.  Learn more> 

Product Release

Inside Dell EMC’s New Soup-to-Nuts Data Science PlatformDell EMC today unveiled a new converged platform that combines the hardware and software data scientists need into a single package. By enabling data scientists to self-provision a set of cultivated Hadoop resources, they can start crunching data and delivering insights from Hadoop in a matter of weeks, rather than the months or years it often takes to stand up a working cluster, the company says. This is Datanami’s review Learn more> 
Published in
0 Comments

Leave a reply

Log in with your credentials

or    

Forgot your details?

Create Account