Machine Learning Hindsight
January 18, 2016
Have you ever found yourself saying, “If I only knew then, what I know now”? It is a moment we all experience, but instead of stewing over our past mistakes it is better to share the lessons we’ve learned with others. Data scientist Peadar Coyle learned some valuable lessons when he first started working with machine learning. He discusses three main things he learned in the article, “Three Things I Wish I Knew Earlier About Machine Learning.”
Here are the three items he wishes he knew then about machine learning, but know now:
- “Getting models into production is a lot more than just micro services
- Feature selection and feature extraction are really hard to learn from a book
- The evaluation phase is really important”
Developing models is an easy step, but putting them in production is difficult. There are many major steps that need attending to and doing all of the little jobs isn’t feasible on huge projects. Peadar recommends outsourcing when you can. Books and online information are good reference tools, but when they cannot be applied to actual situations the knowledge is useless. Paedar learned that real world experience has no comparison. When it comes to testing, it is a very important thing. Very much as real world experience is invaluable, so is the evaluation. Life does not hand perfect datasets for experimentation and testing different situations will better evaluate the model.
Paedar’s advice applies to machine learning, but it applies more to life in general.
Whitney Grace, January 18, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Strong and Loud or Quiet and Weak, Googles Robot Grandkids Fail to Impress the Marines
January 15, 2016
The article titled Why the Marines Don’t Want Google’s Robot Soldiers in Combat on Fortune discusses the downside of the Google-owned company Boston Dynamics’ robots. You might guess, moral concerns, or more realistically, funding. But you would be wrong, since DARPA already shelled out over $30 million for the four-legged battle bots. Instead, the issue is that a single robot, which looks like a huge insect wearing a helmet and knee and elbow pads, emits a noise akin to a motorcycle revving, or a jackhammer drilling, for small movements. The article explains,
“Anyone who’s seen Boston Dynamics’ four-legged robots in action typically is wowed by their speed, strength, and agility, but also note how loud they are. They sound like chainsaws on steroids. And that decibel level is apparently a problem for potential customers, namely the U.S. military.
For Marines who took the robot out for a spin, that noise is apparently a deal breaker. “They took it as it was: a loud robot that’s going to give away their position.”
The reason for all this hullaballoo on the part of the robot is its gas engine, intended for increased robustness. The military was looking for a useful helpmate capable of carrying heavy loads of up to 400 lbs. There has been some back and forth between military representatives and Boston Dynamics, but the current state of affairs seems to be a quieter, and weaker, robot. Not ideal.
Chelsea Kerwin, January 15, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
The Duck Quacks 12 Million Queries
January 14, 2016
DuckDuckGo keeps waddling through its search queries and quacking that it will not track its users information. DuckDuckGo has remained a small search engine, but its privacy services are chipping away at Google and search engines’ user base. TechViral shares that “DuckDuckGo The Anti-Google Search Engine Just Reached A New Milestone” and it is reaching twelve million search queries in one day!
In 2015, DuckDuckGo received 3.25 billion search queries, showing a 74 percent increase compared to the 2014 data. While DuckDuckGo is a private oasis in a sea of tracking cookies, it still uses targeted ads. However, unlike Google DuckDuckGo only uses ads based on the immediate keywords used in a search query and doesn’t store user information. It wipes the search engine clean with each use.
DuckDuckGo’s increase of visitors has attracted partnerships with Mozilla and Apple. The private search engine is a for profit business, but it does have different goals than Google.
“Otherwise, it should be noted that although he refuses to have the same practices as Google, DuckDuckGo already making profits, yes that’s true. And the company’s CEO, Gabriel Weinberg, stop to think it is necessary to collect information about users to monetize a search engine: ‘You type car and you see an advertisement for a car, Google follows you on all these sites because it operates huge advertising networks and other properties. So they need these data for search engines to follow you.’ ”
DuckDuckGo offers a great service for privacy, while it is gaining more users it doesn’t offer the plethora of services Google does. DuckDuckGo, why not try private email, free office programs, and online data storage? Would you still be the same if you offered these services?
Whitney Grace, January 14, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Advice for Marketers, Not Consumers, on the Present and Future States of Location Data Technology
January 14, 2016
The article on Mashable titled Location Data’s Dirty Secret: How Accuracy is Getting Lost in Today’s Data Shuffle relates the bad news for marketers, and hugely relieving news for paranoid consumers, that location data quality is far from precise. The money being funneled into location-targeted mobile ad revenues is only part of the picture, but it does illustrate the potential power of this technology for marketers, who want to know everything they can about shopping habits and habits in general. But they may be spending on useless data. In fact, the article states,
“Studies indicate that more than half of mobile location data is inaccurate. In fact, a report from the MMA offers a laundry list of variables that negatively impact location data quality. Culprits include a “lack of accuracy standards and market education,” “urban density,” “inaccurate interpretations” of location data that have been translated into a latitude/longitude coordinate and poor “data freshness.”
The article is largely optimistic that if marketers do a little research into the source of their locating data, they will know whether it can be trusted or not. That, and an objective third party will help marketers avoid big money-wasting mistakes. Must be nice to be a marketer instead of a consumer, the latter has little chance to avoid being a pawn followed around the chess board by her cell phone.
Chelsea Kerwin, January 14, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Big Data Shows Its Return on Investment
January 13, 2016
Big data was the word that buzzed through the IT community and made companies revaluate their data analytics and consider new ways to use structured and unstructured information to their benefit. Business2Community shares how big data has affected companies in sixteen case studies: “16 Case Studies Of Companies Proving ROI Of Big Data.” One of the problems companies faced when implementing a big data plan was whether or not they would see a return on their investment. Some companies saw an immediate return, but others are still scratching their heads. Enough time has passed to see how various corporations in different industries have leaned.
Companies remain committed to implementing big data plans into their frameworks, most of what they want to derive from big data is how to use it effectively:
- “91% of marketing leaders believe successful brands use customer data to drive business decisions (source: BRITE/NYAMA)
- 87% agree capturing and sharing the right data is important to effectively measuring ROI in their own company (BRITE/NYAMA)
- 86% of people are willing to pay more for a great customer experience with a brand (souce: Lunch Pail)”
General Electric uses big data to test their products’ efficiently and the crunch the analytics to increase productivity. The Weather Channel analyzes its users behavior patterns along with climate data in individual areas to become an advertising warehouse. The big retailer Wal-Mart had added machine learning, synonym mining, and text analysis to increase search result relevancy. Semantic search has also increased online shopping by ten percent.
The article highlights many other big brand companies and how big data has become a boon for businesses looking to increase their customer relations, increase sales, and improve their services.
Whitney Grace, January 13, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
The Business World Is Not Prepared for a Cyber Attack
January 12, 2016
Cyber threats have been a concerning topics since computers became functional and daily tools for people. The idea of a hacker brings up images of IT geeks sitting in a dark basement with their laptops and cracking top secret codes in a matter of keystrokes. Hacking has turned from a limited crime to a huge international problem comparable to the mafia. While hackers are interested in targeting individuals, the bolder thieves target big businesses. News of Bahrain shares that “Biz Not Prepared For Cyber Threat,” translated from headline speech that means the business world would not withstand a cyber attack.
KPMG International released the 2015 KPMG CEO Outlook Study that found businesses are aware of risks associated with cyber attacks, but only forty-nine percent have prepared for one. The study surveyed 1,200 CEOs and one out of five are concerned about cyber risks. The concern has led many CEOs to take action with security measures and safety plans.
“ ‘The most innovative companies have recognized that cyber security is a customer experience, not just a risk that needs to be managed or a line item in the budget. In Bahrain, some firms are finding ways to turn cyber preparedness into a competitive advantage with customers, and they are using this as a differentiator.’ ”
Many companies that are attacked thought they were prepared for any threats, but they underestimated hackers’ intelligence, sophistication, and persistence.
Some of the companies with good cyber security are advertising their technical achievements to prevent attacks. It is a desirable feature, especially as more information is housed on cloud storage and businesses need to be aware of potential threats.
Whitney Grace, January 12, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
There Is a Hole in the Cloud
January 11, 2016
Everyone is running to the cloud to reserve their own personal data spot. Companies have migrated their services to the cloud to serve a growing mobile clientele. If you are not on the cloud, it is like you’re still using an old flip phone. The cloud is a viable and useful service that allows people to access their data anytime and anywhere. Business Insider reveals that cloud usage is heavily concentrated in the US: “Latest Data From The Valley’s Oldest VC Firm Shows One Big Flaw In The Hype Around The Cloud.”
Bessemer Venture Partners is the longest running venture capitalist company in Silicon Valley. To celebrate its 100th cloud investment, it surveyed where the company’s cloud investments are located. Seventy-six of the startups are in the US, eleven are in Israel, and four are in Canada.
“The fact that less than one-quarter of BVP’s cloud investments are in non-US startups shows the adoption of cloud technologies is lagging in the rest of the world. It’s also a reminder that, even after all these years of cloud hype, many countries are still concerned about some aspects of cloud technology.”
Cloud adoption around the world is slow due to the US invents a lot of new technology and the rest of the world must catch up. Security is another big concern and companies are hesitant to store sensitive information on a system with issues.
The cloud has only been on the market for ten years and has only gained attention in the past five. Cell phones, laptops, and using open source software took time to catch on as well.
Whitney Grace, January 11, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
IBMs CFO Reveals IBMs Innovation Strategy: Why Not Ask Watson
January 11, 2016
The article on TechTarget titled IBM CFO Schroeter on the Company’s Innovation Strategy delves into the mind of Martin Schroeter regarding IBM’s strategy for chasing innovation in healthcare and big data. This year alone IBM acquired three healthcare companies with data on roughly one hundred million people as well as massive amounts of data on medical conditions. Additionally, as the article relates,
“IBM’s purchase of The Weather Co.’s data processing and analytics operations brought the company a “massive ingestion machine,” which plays straight into its IoT strategy, Schroeter said. The ingestion system pulls in 4 GB of data per second, he said, and runs a lot of analytics as users generate weather forecasts for their geographies. The Weather Co. system will be the basis for the company’s Internet of Things platform, he said.”
One of many interesting tidbits from the mouth of Schroeter was this gem about companies being willing to “disrupt [themselves]” to ensure updated and long-term strategies that align technological advancement with business development. The hurtling pace of technology has even meant IBM coming up with a predictive system to speed up the due diligence process during acquisitions. What once took weeks to analyze and often lost IBM deals has now been streamlined to a single day’s work. Kaboom.
Chelsea Kerwin, January 11, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
The Long Goodbye of Internet Freedom Heralded by CISA
January 8, 2016
The article on MotherBoard titled Internet Freedom Is Actively Dissolving in America paints a bleak picture of our access to the “open internet.” In spite of the net neutrality win this year, broadband adoption is decreasing, and the number of poor Americans forced to choose between broadband and smartphone internet is on the rise. In addition to these unfortunate trends,
“Congress and President Obama made the Cybersecurity Information Sharing Act a law by including it in a massive budget bill (as an extra gift, Congress stripped away some of the few privacy provisions in what many civil liberties groups are calling a “surveillance bill”)… Finally, the FBI and NSA have taken strong stands against encryption, one of the few ways that activists, journalists, regular citizens, and yes, criminals and terrorists can communicate with each other without the government spying.”
What this means for search and for our access to the Internet in general, is yet to be seen. The effects of security laws and encryption opposition will obviously be far-reaching, but at what point do we stop getting the information that we need to be informed citizens?
And when you search, if it is not findable, does the information exist?
Chelsea Kerwin, January 8, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
In Scientific Study Hierarchy Is Observed and Found Problematic to Cooperation
January 8, 2016
The article titled Hierarchy is Detrimental for Human Cooperation on Nature.Com Scientific Reports discusses the findings of scientists related to social dynamics in human behavior. The abstract explains in no uncertain terms that hierarchies cause problems among human groups. Perhaps surprisingly to many millennials, hierarchies actually forestall cooperation. The article explains the circumstances of the study,
“Participants competed to earn hierarchy positions and then could cooperate with another individual in the hierarchy by investing in a common effort. Cooperation was achieved if the combined investments exceeded a threshold, and the higher ranked individual distributed the spoils unless control was contested by the partner. Compared to a condition lacking hierarchy, cooperation declined in the presence of a hierarchy due to a decrease in investment by lower ranked individuals.”
The study goes on to explain that regardless of whether power or rank was earned or arbitrary (think boss vs. boss’s son), it was “detrimental to cooperation.” It also goes into great detail on how to achieve superior cooperation through partnership and without an underlying hierarchical structure. There are lessons to take away from this study in the many fields, and the article is mainly focused on economic metaphors, but what about search vendors? Organization does, after all, have value.
Chelsea Kerwin, January 8, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
-
- Subscribe to Beyond Search
Feature archive
News archive
-
