November 09, 2015

H2O and the Magic of Machine Learning

What’s really interesting to me is if you think about how this pattern recognition technology can enable insights to be derived from data. It’s really compelling because before machine learning you needed people to come up with creative ideas and essentially superimpose their own perspective on to a data set. Whereas now with machine learning, you can, essentially, get some insights via machine learning into the different patterns that exist in the data. You’re basically facilitating that first step of a human being working with the information to try to understand what the segments are, maybe what they mean and give some direction to where that person goes and try to better analyze the data and then come up with some ideas, right?

Watch a man turn his arm into a virtual keyboard using a device

Japanese electronics powerhouse NEC is working on a system called, fittingly, “ARmKeypad,” which creates a virtual keyboard using a set of glasses and a smartwatch, The Wall Street Journal reports. NEC told the Wall Street Journal the keyboard’s main advantage is that, unlike voice-operated devices, it can be operated in noisy environments. The company sees it being useful in healthcare, manufacturing, document management, and security. The idea of typing on whatever surface we happen to be looking at is something long promised by science fiction films but that's been slower to move into reality. And the arm is a start. The company plans to publicly release the ARmKeypad in 2016.

Gartner Predicts Our Digital Future

Here’s a scene from our digital future: You sit down to dinner at a restaurant where your server was selected by a “robo-boss” based on an optimized match of personality and interaction profile, and the angle at which he presents your plate, or how quickly he smiles can be evaluated for further review. Or, perhaps you walk into a store to try on clothes and ask the digital customer assistant embedded in the mirror to recommend an outfit in your size, in stock and on sale. Afterwards, you simply tell it to bill you from your mobile and skip the checkout line. These scenarios describe two predictions in what will be an algorithmic and smart machine driven world where people and machines must define harmonious relationships.

Deep Learning in a Nutshell: Core Concepts

Feature engineering is the most important skill when you want to achieve good results for most predictions tasks. However, it is difficult to learn and master since different data sets and different kinds of data require different feature engineering approaches. Only crude guidelines exist, which makes feature engineering more of an art than a science. Features that are usable for one data set often are not usable for other data sets (for example the next image data set only contains land animals). The difficulty of feature engineering and the effort involved is the main reason to seek algorithms that can learn features; that is, algorithms that automatically engineer features.

Taming today's cyberthreat landscape: A CIO checklist

One of the few things the "experts" seem to agree upon is that cybercrime is a clear and present danger to our national security. These issues have gone way beyond the province of esoteric IT journals and cultish science fiction novels -- they have invaded our daily collective consciousness and well-being as individuals, as families, as companies, as governments, as a society and as a culture at large. Many opine at great length on how the cyber landscape has become the new battleground upon which future wars will be fought: Nations will rise and fall based upon their techno-prowess to aggressively attack and defend against the new breed of cybercriminals.

Flash storage: Is there a tipping point anytime soon?

Flash is still too expensive for general use when compared to conventional disk storage, which offers much higher capacities than flash at a fraction of the cost. Now the signs are that we are going to see that situation changing -- but if the analysts are to be believed, that is not going to happen quickly. "When you go to refresh your technology, you look at flash and now you think about phasing it in," said Valdis Filks, the expert on flash storage at analyst house Gartner. But as he explained, this will take some time: "It is only after two or three cycles [of technology upgrades] that most companies can bring in a complete change in technology, like a move to all flash".

The Data Movement Dilemma

Most organizations quickly realize, if they analyze their situation, that it isn’t possible to adopt a single solution for data movement. There will already be a variety of different approaches implemented throughout their systems, and in most cases, the technology will be fit for purpose. Nevertheless, they are also likely to discover that the dynamics of new applications, coupled with data growth in many areas, will inevitably stress some data movement technologies as time passes. The problem is to “maintain the plumbing” and replace only those parts that are under stress and in danger of being overwhelmed. The only solution that comes to mind is a capability that can to manage data movement with an automation capability, like Automic’s, that can monitor every data transfer and provide actionable information when specific service levels are threatened.

Japan its own enemy in push to improve cybersecurity

"In the U.S., if they find a problem, they have to report," he said. "The Japanese engineer feels he fails his duty if he escalates a report. They feel ashamed." To be sure, the cybersecurity industry around the world, not just in Japan, frequently echoes the call for greater transparency within and among organizations. The U.S. Senate last month passed the Cybersecurity Information Sharing Act to ease data sharing between private companies and the government for security purposes, although civil liberties advocates warned it posed a threat to privacy. But the problem may be particularly acute for Japan's private sector behemoths and government ministries. These sprawling bureaucracies are wrapped in a "negative culture that cuts against wanting to communicate quickly," said William H. Saito, the top cybersecurity adviser to Prime Minister Shinzo Abe.

CloudFlare Supplies Security At Network's Edge

"We're equivalent to CloudFront, Amazon's Edge product, but we're built to have more flexibility," said Matthew Prince, cofounder and CEO of CloudFlare in San Francisco. Prince was literally moving between 665 Third Street and 101 Townsend Street, CloudFlare's new home, when InformationWeek caught up with him in the South of Market section of the city, not far from where the Giants baseball team plays at AT&T Stadium. CloudFlare's building is still a work in progress -- a former warehouse converted to offices a long time ago but still needing a lot of modernization and some finishing touches. The move was delayed by the need to get the building rezoned for offices.

How Carders Can Use eBay as a Virtual ATM

So-called “triangulation fraud” — scammers using stolen cards to buy merchandise won at auction by other eBay members — is not a new scam. But it’s a crime that’s getting more sophisticated and automated, at least according to a victim retailer who reached out to KrebsOnSecurity recently after he was walloped in one such fraud scheme. The victim company — which spoke on condition of anonymity — has a fairly strong e-commerce presence, and is growing rapidly. For the past two years, it was among the Top 500 online retailers as ranked by The company was hit with over 40 orders across three weeks for products that later traced back to stolen credit card data.

Quote for the day:

"The leader who exercises power with honor will work from the inside out, starting with himself." --Blaine Lee