Data Data Data

Tuesday, December 29, 2015

In 2015, APIs for IBM's Watson system were front and center as a means to bring cognitive computing applications to a broader corporate audience. "People don't necessarily want to buy a million-dollar system to run Watson," IDC's David Schubmehl said. "But PaaS is well-suited for cognitive platforms. People can use Bluemix services and start working with one or two APIs, rather than use the whole system." He added that IBM's March 2015 purchase of AlchemyAPI -- a deep-learning AI technology startup -- was also notable, as it brought to Big Blue a popular set of developer APIs that can help Watson in areas beyond machine learning applications. READ IBM Watson APIs hold key to broader cognitive computing use.

Author Medina describes cybernetics experiment in 1970s Chile

Does a 1970s Utopian technology effort offer useful guides for those trying to assess the progress of new technology today? In one case, at least, yes. It is the story of Salvador Allende's attempt to build a working Socialist government in Chile with computer cybernetics.

The tale is told especially well, under the able hands of author and researcher Eden Medina. Medina rolls up the takeaways in a recent article in Jacobin magazine. It is a summary of some important lessons garnered during work on her 2013 book, The Cybernetic Revolutionaries. (Prose continues here.)

Saturday, December 26, 2015

TensorFlow makes news in 2015

On the face of it, it would appear that TensorFlow received an inordinate amount of attention in 015 for just a machine learning engine. But publicity works that way. Google is a $66 billion-per-year company, and buzz automatically goes with that. But a series of pieces by Cade Metz in Wired were fairly illuminating.

You push the little valve down, and the music goes around, and it comes out.

It seems that Google open sourcing, to some extent, TensorFlow gave it some lift. My take would be that they would like others to write the Java and JavaScript Notebooks, and create long lists of libraries, to give it the panache of Apache Spark, which is provably hot, due to its metered Apache klingonage. Is it an attempt to take wind out of Spark's sails? Some people say that Spark is general-purpose, and thus not as good for machine learning as is TensorFlow, which has no other use in life but to do machine learning.

Yet another Wired piece by Metz discusses the upswing in use of GPUs for machine learning. One banner that TensorFlow seems to forward is the use of GPUs for machine learning. It seems to be a growing meme.

One wag opines that TensorFlow smells like something that might work autonomous car data. In the back of my mind I can hear the words of someone who told me the difference between IBM and Watson and Google and TensorFlow is that the former is about software to enable enterprises to make consumer products, while the latter is about making consumer products. [Just last week, Google said it would launch such a vehicle with Ford. Which could play to the idea that it doesnt want to make cars, it wants to get data on drivers.]

As described Metz's article deep learning is the same as machine learning. Metz points out that the new goop in the secret sauce is the increase in both available processing and available data – suggesting that the algorithms are not dramatically different than in the past. [Although the author goes on to say the algorithms are evolving, and that gifted individuals are behind that evolution.]

Author Metz and source Lukas Biewald of Crowdflower note that Google open sourced the machine learning software, but not the data. Others criticize the fact that, while you can run Tensorflow on your own machine (that could include a GPU board), they are keeping the distributed version to themselves.

Links
https://www.youtube.com/watch?v=ENZoY4mLgDE
http://www.wired.com/2015/11/google-open-sources-its-artificial-intelligence-engine/
http://www.kdnuggets.com/2015/11/google-tensorflow-deep-learning-disappoints.html
http://www.wired.com/2015/11/google-open-sourcing-tensorflow-shows-ais-future-is-data-not-code/

http://www.wired.com/2015/11/googles-open-source-ai-tensorflow-signals-fast-changing-hardware-world/
https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi-5K0GDYQsrw_oRhC_2hQ0lOp5xC0n1GRxtnpViaFNmriieL3FWkHQRf9QJS_5W7ZMPilxA-MxibJO4A9dygaRSqj0x15klMEstjZ7JDbVtg282JnG8IRd4wTTBEfS828paM2UEQi-Vk4/s1600/cifar10_2.gif
http://www.tensorflow.org/
http://googleresearch.blogspot.com/2015/11/tensorflow-googles-latest-machine_9.html

http://googleresearch.blogspot.com/2015/11/computer-respond-to-this-email.html
https://www.youtube.com/watch?v=46Jzu-xWIBk
https://www.youtube.com/watch?v=gY9DewL6Dqk
http://blogs.nvidia.com/blog/2015/03/18/google-gpu/

Wednesday, December 2, 2015

Data Decisions

Let's not paint big data too darkly. It brings hope -- not just of making money, but of improving social institutions, helping to cure disease and more. However, hope can be accompanied by fear. While privacy is the chief public concern in the new world of voluminous information, there are others, as well. The chance that bad decisions might be made based on misreading big data is one of them. - Jack Vaughan

read more at http://searchdatamanagement.techtarget.com/news/2240185076/Business-decision-making-must-progress-in-the-age-of-big-data

Wednesday, November 25, 2015

Talking Data Podcast ponders analytical algorithms to help save whales

A right whale and her right calf. Source: NOAA

In a recent Talking Data podcast guest Kristen Khan of NOAA said she and her colleagues have watched the growth of machine learning projects that identify images. They have wondered: If such analytical algorithms could trim time from what is now a very labor intensive whale identification task, could that free up staff for more proactive efforts to save the whales? A contest sponsored by MathWorks is trying to find the answer. Listen here.

Tuesday, November 17, 2015

Architectures from different views

rdf..its all about the triples.. you nurse it an rehearse it. the idea of a sentence.

take a bunch of different types of data from different sources .. put them all together in triples

on another level you layer it on a physical data infrastructure architecture

Monday, November 9, 2015

Tensor

Let’s look inside a Learning Machine - Google Tensor

https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi-5K0GDYQsrw_oRhC_2hQ0lOp5xC0n1GRxtnpViaFNmriieL3FWkHQRf9QJS_5W7ZMPilxA-MxibJO4A9dygaRSqj0x15klMEstjZ7JDbVtg282JnG8IRd4wTTBEfS828paM2UEQi-Vk4/s1600/cifar10_2.gif
http://www.tensorflow.org/
http://googleresearch.blogspot.com/2015/11/tensorflow-googles-latest-machine_9.html
http://googleresearch.blogspot.com/2015/11/computer-respond-to-this-email.html
https://www.youtube.com/watch?v=46Jzu-xWIBk
https://www.youtube.com/watch?v=gY9DewL6Dqk

Friday, November 6, 2015

Machine learning for better medical income on TAP

Michael Draugelis, is chief data scientist, The University of Pennsylvania, Penn Medicine. He came to this gig in a way roundabout. You see, his wife went into shock while giving birth to their child, Chubsy Ubsy. Mother and child are doing well, but the experience made Draugelis wonder. That's because of his background in US Missile Defense Agency, where they did a lot of work about forecasting clues to impending events. He appeared at Strata + East to discuss all this.

When he got to Penn Med he set out to focus on Sepsis, an unfortunately leading cause of death for people who go to the hospital to fix something else. His data scientist efforts revolve around something called Penn Signals. "A nerd wonderland," says he. The point is to use evidence-based computing to see who is really really at risk. Right now!

Now, of course, in reality, he is doing what all the machine learning people are doing:

We have some early proof of concepts that have created new signals..that we are feeding back into our algorithms.

What is next?

I hope to move on from prediction to auto reasoning.

That is where his work with Intel comes in.

The company, which would like to fight disease just as much as it would like to sell chips that go into servers on Web farms, has concluded that big data development, where egg headed data scientists create models in Python etc., that must be thrown over the wall, and rejiggered in Java by 'real' developers, and then sent back to fix, and then back over the wall again etc., was just going to take too much time. Why not more of a cloud development paradigm? Based on open source? Tested by Intel and the like of Penn? It's called TAP, for Trusted Analytics Platform.

Says our man Draugelis:

My data scientists need an environment that they can build quickly, select their analytic tools at scale, and a platform that can support it. We have been excited to work with Intel to explore this new open source project called TAP.

As a colleague said, or more precisely asked: " What was the last big software announcement Intel made? And what came of it? And yes the answer is The Intel Hadoop Distribution and what came of it was a ceding of the work to Cloudera.So to coin a phrase: Time will tell. - Jack Vaughan

For more

Go to github to find out what it would be like to be a cloud developer these days.

Listen to a podcast where we talk with Intel's Vin Sharma about TAP

You see that embedded video above, this here is the linke to it, like actually.