Artificial Intelligence Revenue Forecast at $36.8b Issue 69: CognitionX Data Science, AI and Machine Learning Briefing


So many conferences, Big Data is big in October and November. Come and join us at:

Big Data in Use Conference at Big Data Week London 2016 – Apply with this code: CogX25 to get a 25% discount

Big Data LDN – We are hosting a MeetUp bringing Deep Learning to BigData LDN

LT Accelerate – Get a 20% discount (code: PCognitionX) to Europe’s top NLP event 


Tabitha ‘BigLanyard’ Goldstaub

Impact on the economy

5G in the UK 2018 and UCL getting more funding into the space

AI is not only a key enabler for how 5G will function, but important for new applications that will benefit from the future networks. Niall Roche UCL’s Enterprise Tech Lead gives us more insight on this in his CognitionX blog.

Ericsson has now announced that it has concluded a 5G field trial with Swedish telecom Telia that saw the transfer of information at a rate 40 times faster than what’s currently achievable via 4G

And also announces that UCL will help coordinate an Open Call for companies to apply for funding in the region of £25,000 for projects using elements of 5G that can make a significant impact to their products or services. The first TRIANGLE Open call is open until October 31st for more details visit the blog.

Audio for the journey home

We need to talk about the kids

This episode of the Getting Smart podcast invites software engineer Gerald Huff to talk about role that Artificial Intelligence can play in education. Links to the some of the papers and reports

Feed your mind over lunch

Data science in a cocoon doesn’t help anyone

TechTarget nicely describes the opinion of Ritika Gunnar, a vice president in IBM’s data of Data Science: 

“The data science soul… understands the nuances of the business. Looking for trends is just database stuff and SQL queries….data science is more about understating the outliers and finding patterns that were not evident. You’ve got to dig much deeper.”

As one of our partners said on Wednesday it’s about finding the pearl not showing hundreds of oysters.

Pure, unadulterated research

Emoji me what could I mean?
In this paper Isabelle Augenstein and team release emoji2vec, pre-trained embeddings for all Unicode emoji which are learned from their description in the Unicode emoji standard. The resulting emoji embeddings can be readily used in downstream social natural language processing applications alongside word2vec.

I’m excited to see a sudden improvement in Twitter sentiment analysis models after this. Maybe even in their valuation?!

Data Visualisation

NY and it’s Taxi Cab problem

MAPD presents a staggering, 1.2 billion row taxi/limo/uber/lyft dataset from New York. Processing the data in such amazing ways relies on nearly 40,000 GPU cores and Low Level Virtual Machines (which are elaborated upon here)

Infographic: 67 startups making your home smarter
For today’s CB insights fix – this infographic shows the pervasive extent to which AI is in the home.

Education, training and advice we rate

Recurrent Neural Network Tutorial #4 – Implementing a GRU/LSTM RNN with Python and Theano
In this post ODC teach us about LSTM (Long Short Term Memory) networks and GRUs (Gated Recurrent Units).

If you inherently understand or are excited by this post, please get in touch, we have a client looking for someone with Python & Theano skills.

Business impact of AI

AI Top 20 use-cases for 2025
According to a report from The AI for Enterprise conference the top use-cases in 2025 will be:

Algorithmic trading at $2.4 billion
Image recognition, classification, and tagging $2.3 billion
Efficient, scalable processing of patient data at $2.3 billion.

Check out the infographic below (NB: you’ll need to give them your email to see larger)

Copyright © 2016 CognitionX, All rights reserved.


Published in

Leave a reply

Log in with your credentials


Forgot your details?

Create Account