Deeper neural nets often yield harder optimization problems. At the heart of deep learning lies a hard optimization problem. So hard that for several decades after the introduction of neural networks, the difficulty of optimization on deep neural networks was a barrier to their mainstream usage and contributed to their decline in the 1990s and 2000s. Since then, we have overcome this issue. In this post, I explore the "hardness" in optimizing neural networks and see what the theory has to say.

This month, Portugal is descended upon not by tourists in search of sunshine but by some estimated 55,000 tech-savvy individuals to attend an event quoted by Bloomberg as'Davos for Geeks'. I am of course talking about Web-Summit. For those unknown about Web-Summit, it's Europe's largest Technology Marketplace and invites companies within the tech industry to share their knowledge, and inspire through talks on how to solve some of the world's most pressing issues through new cutting edge systems and intelligent innovation. Whilst watching these talks one message resonated louder and clearer than all others – the future of technology is accelerating at an alarming rate, leaving all others who do not keep up the pace by the wayside. This message didn't, however, alarm me but rather triggered hope that we are advancing towards a technological level where we can tackle vital, real world issues such as: Climate change – our planet is warming causing problems such as rising sea levels and the first ever climate refugees.

Artificial Intelligence (AI) is the ability for a machine to solve problems by learning over time. It is similar to the way a brain processes information when making a decision. The term was coined by John McCarthy in 1956 who also quipped, "as soon as it works, no one calls it AI anymore." The technology is now so pervasive we no longer associate it with science fiction. From protecting you against credit card fraud, to recommending shows on Netflix and self-driving cars, this type of intelligence is spreading.

The Ishikawa Watanabe Laboratory, the University of Tokyo laboratory has all rights to the materials shown in the video. The paper "Learning to Simplify: Fully Convolutional Networks for Rough Sketch Cleanup" and its online demo is available here: http://hi.cs.waseda.ac.jp/ esimo/en/r... http://hi.cs.waseda.ac.jp:8081/ Recommended for you: Rocking Out With Convolutions - https://www.youtube.com/watch?v JKYQO... Separable Subsurface Scattering - https://www.youtube.com/watch?v 72_iA... WaveNet by Google DeepMind - https://www.youtube.com/watch?v CqFIV... WE WOULD LIKE TO THANK OUR GENEROUS PATREON SUPPORTERS WHO MAKE TWO MINUTE PAPERS POSSIBLE: Sunil Kim, Julian Josephs, Daniel John Benton, Dave Rushton-Smith, Benjamin Kang. Subscribe if you would like to see more of these! - http://www.youtube.com/subscription_c... Image credits: Bitmap and vector images (two of them): Wikipedia - https://en.wikipedia.org/wiki/Vector_... and https://en.wikipedia.org/wiki/Image_t... Image resolution: Wikipedia - https://en.wikipedia.org/wiki/Image_r... Vectorization: Wikipedia - https://en.wikipedia.org/wiki/Image_t... Thumbnail background - https://pixabay.com/photo-1281718/ Music: Dat Groove by Audionautix is licensed under a Creative Commons Attribution license (https://creativecommons.org/licenses/...) Artist: http://audionautix.com/

Artificial intelligence sounds mysterious to many people, but it has been a part of our life, an expert said at the Third World Internet Conference in Wuzhen, East China's Zhejiang province, on Wednesday. The voice recognition function for mobile internet search used by Baidu, for instance, is an application of artificial intelligence, said Sun Ninghui, director of the Institute of Computing Technology, Chinese Academy of Sciences. The tech company's "Baidu Doctor", which can simulate dialogues between a patient and a doctor, and read large amount of literature and the patient' records, is another example of how artificial intelligence is applied, he added. Besides, artificial intelligence is also used in image recognition and teaching, according to Sun. More applications such as autonomous vehicles would soon enter our life, Sun said, referring to Baidu's driverless vehicle which recently underwent road tests.

Bad hair days may soon be prevented by just looking at your smart phone. Parham Aarabi and Wenzhi Guo, researchers at the University of Toronto Faculty of Applied Science & Engineering, have developed a machine learning algorithm that learns directly from human instructions rather than an existing set of examples. The algorithm outperformed conventional methods of training neural networks by 160 percent and outperformed its own training by 9 percent. The algorithm is seen as a significant leap forward for artificial intelligence as it learned to recognize hair in pictures with greater reliability than that enabled by the training. The two researchers were able to train the algorithm to identify people's hair in photographs--a much more difficult task for computers than it is for humans.

The year 1995 was heralded as the beginning of the "New Economy." Digital communication was set to upend markets and change everything. It wasn't that we didn't recognize that something changed. It was that we recognized that the old economics lens remained useful for looking at the changes taking place. The economics of the "New Economy" could be described at a high level: Digital technology would cause a reduction in the cost of search and communication.

Infosys has made an investment of R14.5 crore in a Denmark-based artificial intelligence start-up, Unsilo. Founded in 2012, this Danish company is focused on solutions in the area of advanced text analysis. Infosys has earmarked $500 million for its innovation fund which invests in start-ups across the globe. Of this, $250 million has been set aside for start-ups in India. "We will partner with Unsilo to bring their artificial intelligence and machine learning technology to our global clients. They join and expanding portfolio of innovative young companies from around the world that Infosys works with to help enterprises drive their digital transformation," said Ritika Suri, executive vice president & global head of corporate development & ventures at Infosys.

Want to play around with Google's AI projects? Well, now you can thanks to a website which makes the company's tinkering with artificial intelligence available for the public to sample. If you head over to the AI Experiments web page, you'll find eight different experiments on offer, designed to illustrate what machine learning can do and appeal to a wider audience (with the option to submit your own work, to boot, if you've concocted anything along these lines yourself). The web-based apps include one which is an experiment in seeing whether a neural network can recognise a sketch. With Quick Draw, you simply doodle something on your touchscreen and the AI gives its best guess as to what that object is.