Author Archives for joanfihu_adl

Spark promises to up-end Hadoop, but in a good way

Date: February 12, 2015 Published by Leave your thoughts

Not only can Spark work with data in-memory, making queries 100x faster than MapReduce, but Spark queries on disk also run 10x faster. One of the biggest problems with big data is that the technology is either insanely expensive, insanely complicated, or both. This is why traditional Hadoop powerhouses like Cloudera have maintained some commitment to MapReduce as they’ve dramatically increased their commitment to Spark.... View Article


The Upside of Artificial Intelligence Development

Date: February 11, 2015 Published by Leave your thoughts

Derrick Harris (@derrickharris), a senior writer at Gigaom, asserts that the fact of the matter is that artificial intelligence (at least the narrow kind) is here, is real, and is getting better. [“Artificial intelligence is real now and it’s just getting started,” Gigaom, 9 January 2015] He explains: Will artificial intelligence be disruptive? Timothy B. Lee (@binarybits), senior editor at @voxdotcom, also believes that the... View Article


Where’s Waldo ruined by a doctoral student and machine learning

Date: February 10, 2015 Published by Leave your thoughts

All those hours you spent scouring Where’s Waldo books as a child were wasted now that a doctoral student at Michigan State University’s High-Performance Computing Center has used a machine to speed the process up. Luckily, Olson was able to get his hands on the coordinate positions of all 68 Waldo locations from the seven primary editions of the Where’s Waldo books. Rather than loaf... View Article


Google Brain’s Co-Inventor Tells Why He’s Building Chinese Neural Networks

Date: February 10, 2015 Published by Leave your thoughts

Deep learning algorithms are very good at one thing today: learning input and mapping it to an output. Ng and I chatted about the challenges he faces leading the efforts for “China’s Google” to understand our world through deep learning. A single “neuron” in a neural network is an incredibly simple mathematical function that captures a minuscule fraction of the complexity of a biological neuron.... View Article


Recommending music on Spotify with deep learning

Date: February 6, 2015 Published by Leave your thoughts

There is quite a large semantic gap between music audio on the one hand, and the various aspects of music that affect listener preferences on the other hand. Traditionally, Spotify has relied mostly on collaborative filtering approaches to power their recommendations. The idea of collaborative filtering is to determine the users’ preferences from historical usage data. Collaborative filtering algorithms will not pick up on this.... View Article


‘Soft’ Artificial Intelligence Is Suddenly Everywhere

Date: February 5, 2015 Published by Leave your thoughts

Human intelligence has evolved over millions of years. Our highly complex IT systems must become much more autonomic and resilient, capable of self-healing when failures occur and self-protecting when attacked. What really concerns me are the highly complex IT systems that we’re increasingly dependent on in our every day life. We are increasingly developing highly complex socio-technical systems in areas like health care, education, government and cities. These complex systems are composed of many different... View Article


​Facebook open sources AI tools, possibly turbo charges deep learning

Date: February 4, 2015 Published by Leave your thoughts

The social networking giant has a long open source contribution history and has even released documentation on its data center designs and building blocks. Summary:The plain English takeaway is that faster training of neural networks will now be widely available via open source project Torch. By contributing its optimized deep-learning modules and other tools, Facebook could turbo charge numerics, machine learning and computer vision. Facebook... View Article


Google on the Technical Debt of Machine Learning

Date: February 3, 2015 Published by Leave your thoughts

A number of Google researchers and engineers presented their view on the technical debt of using machine learning at the Software Engineering for Machine Learning workshop, which was part of the annual NIPS conference held in Montreal. They identified different aspects of technical debt and came to the conclusion that without proper care, using machine learning or complex data analysis in your company can lead... View Article


Artificial-Intelligence Experts to Explore Turing Test Triathlon

Date: February 2, 2015 Published by Leave your thoughts

The result is a daylong workshop, “Beyond the Turing Test,” where attendees aim to work out an alternative to the current test. That’s essentially the problem that developed over the years with the Loebner Prize version of the Turing test. One of the criticisms of the Turing test is that it emphasizes language at the expense of other skills, such as vision or movement. One... View Article