FutureVigil (tm) » data mining

Web Decision Making-Hunch.com and Caterina Fake (Flickr)

Jeffrey Newman — Sun, 08 Aug 2010 18:43:51 +0000

After exploring some recent efforts at data mining in medicine in this blog, I yeared for more info about the use of data mining in social media generally. This isn’t easy to come by if you are not in this arena. I did find some relevant writings on a fairly new site called Hunch and its co-founder Caterina Fake, who has focused on helping people find what they want on the net, essentially using crowd souring to make better decisions fast. This is a pregant field. Making swifter more precise and better decisions based on individualized wants and interests is a critical function, not just in consumerism but in all business. It is also essential in medicine. So, Hunch has captured my interest. Caterina Fake/Hunch first gets people talking about themselves, their opinions, tastes, beliefs, idiosynchrasies and once they have sufficient data they mine it for correlations to provide cutomized recommendations for the user. (the correlation search seems similar to that being done by Andre Brin and his wife at 123&me–see my recent blogs inculding the data mining of blood data from prior clinical studies for therapeutics in Parkinson’s). Hunch learns interesting things about us. For example people who believe in alien abductions are more likely to drink Pepsi.(okay what else?) People who eat fresh fruit more likely to want to buy the Canon’s EOS 7D camera. On her site, Caterina describes the site as a decision making site customized for you. The Hunch site explains that researchers have establlished that decisions made by diverse and independent groups are superior to decisions made by individuals, including experts. The reason is that knowledge is spread among many people. In choosing what to ask us, Hunch’s question selection algorith (MIT developed), tries to ask questions which will help optimize and rank recommendation outcomes. It creates decisions trees.It is the people who answer questions that train the algorithm proper training to make nuanced recommendations. Now I am beginning to understand the concept of machine learning and searching among data for correlations you might not be able to find without the computer. I’m not sure how the Hunch algorithym knows what questions to ask to get at the important intuitive info but I am on the hunt for this. I am also about to try out Hunch to seek what areas of of strength it has. My goal is to keep an eye on companies like this for the lateral benefits they might have towards the development of medical therapeutics. If you are in this field, please comment. There is also an article in WIRED Mag. on Hunch but it is quite general. There are things I really like about the goals of this company–first–Hunch wants to establish via proof that it can do better in this process than other companies. For example Match.com uses the process to connect people. If the company evolves, it should be able to do so but better. Same with Amazon and books. From what I see, Hunch does not try to tell people what they want but rather to help to define within what they want how to understand what might be better for them based on those wants/needs. Perhaps it will expand to help people define wants and needs for a better more fruitful life.

Sergey Brin’s Search For a Parkinson’s Cure

Jeffrey Newman — Sun, 04 Jul 2010 17:58:48 +0000

http://www.wired.com/wiredscience/2009/03/science-as-sear/

This WIRED article, about Sergey Brin’s use of data-mining and social networking for the advancement of medical science is really worth reading for so many reasons. It explores the interface between data mining and medicine; the slowness of present medical research under traditional paradigms and connects social networking to medical research and more. Brin, co-founder of Google is married to a genetic researcher who founded 23andme http:www.23andme.com a unique DNA profiling company which gathers information from millions of volunteers via created social networks. They send out kits and surveys and the DNA and data sent back is entered into a massive databank from which computer data analysis programs are used to find correlations and trends which might help scientists save time in their work. The work, while rudimentary now, opens up enormous potential for understanding the mechanism of action of various disorders and the genetic basis and potential environmental and lifestyle mitigators for individuals. It’s short term benefits might be most useful to translational medicine experts, whose function is to advance medicine from the lab to the patient, speeding the time frame for actual therapeutics. People in this field create bridges from theory to practice and are attuned to the benefits of trend analysis and thinking in non-traditional ways. The data-mining may also be of benefit to basic science and understanding of cell function and growth, which itself should not be diminished as basic science and traditional medical research and translational medicine are all necessary components to disease cure. However, the WIRED article uses aspirin an excellent example in demonstrating how many major advances in medicine are discovered by accident. Invented in the early 20th century, aspirin was mainly a pain reliever and it was used for decades before its benefits for heart attack prevention were discovered by mistake in the 1960’s and early 70’s. Now, it is a key part of the medical arsenal against cardiovascular disease. There are many such examples and Brin’s thought is that by using data-mining for swift analysis of connections and trends in the medical histories, responses to drugs, urine and sputum samples and surveys of hundreds of thousands of volunteers, we can truncate the time within which such discoveries are made. The surface is just being scratched. I have a good friend who is a scientist and researcher with his own lab. I sent him the article and asked his thoughts. Stay tuned.

Virtual Reality Computer Models Accurately Predict Results of Real Battles

Jeffrey Newman — Sun, 09 May 2010 16:58:41 +0000

New computer modeling based on human behaviors in the past have resulted in the ability of military analysts to predict the probable outcome of particular battles. The research is being done at The University of Maryland and is expected to enhance the battlefield decision making processes in a major way.

What happens if political gridlock kills healthcare

Jeffrey Newman — Tue, 09 Mar 2010 01:13:12 +0000

Few have projected the details of what will occur if the current healthcare initiatives are stalemated in Washington. There has been some general talk of costly overruns for Medicare and a significant increase in interest rates as well as a depressed stock market but these hallmarks really don’t mean much. I would like someone to run an extrapolation as to when the costly healthcare system will begin to depress the economy and the details of what specifically will happen with Medicare. This shouldn’t that hard to do for the Office of The General Budget or the Auditor. If The American Public could see the numbers in black and white and if they were presented with a realistic healthcare proposal with the specifics on how it will be paid for–all set forth in understandable detail, I don’t think there would be as much opposition to a plan as here is now. Of course, I could be wrong.

THE HACK ATTACK ON GOOGLE-WE SEE THE FUTURE

Jeffrey Newman — Tue, 19 Jan 2010 08:36:55 +0000

The hackers which penetrated Google’s defenses recently demonstrated a level of sophistication which paints a frightening picture of just how vulnerable the internet is. The hackers used a combination of encryption, stealth programing and found an unknown hole in internet Explorer. A dozen or more pieces of malware and numerous lawyers of encryption allowed the hackers to bore deeply into company networks and hide their presence as stealthy aliens. The hackers stole intellectual property and gained access to the Gmail accounts of human rights activists and the attacks originated from China. The attacks which targeted 34 companies in the technology, financial and defense sectors was dubbed “Operation Aurora” as this was the name the hackers used for their mission and the name was on the computer of one of the hackers. The initial piece of code was shell code encrypted three times and then was activatedt to penetrate. One malicious code opened a remote backdoor to the computer established by an encrypted secret shannel hiding as an SSL connection to avoid detection. This allowed ongoing access to a computer which afforded the hackers entry to other parts of the network in order to search for login credentials and other data.

Google: Glimmer of a Soul?

Jeffrey Newman — Wed, 13 Jan 2010 15:36:58 +0000

Following on the heels of Google’s announcement that it may pull out of China, Joe Schoendorf, a Partner at Accel Partners, a Silicon Valley venture capital firm told the New York Times “Google has Microsoft on the ropes, and China is arguably the world’s most important market outside of the U.S. You don’t walk away from that on principle.” Mr. Schoendorf may be missing a critical point in long term corporate strategy. If Google is bold enough to take a stand now and close its operations in China, it will serve the establish the company as a leader on the world stage, placing it ahead of any other company in helping to formulate policies dealing with human rights and other legal issues. China, with its expanding role as a global economic power, has had few challenges by corporations or government intervention. Here we have the first glimmer of Google’s soul–what the company is at its heart. This will not be lost on consumers long term. Even publicly traded corporations cannot make decisions based solely on the best interests of its stockholders. Google would go a long way in establishing itself as the visionary company of our time if it pulls out of China. Other companies would do well to watch and do the same. The global arena of nations and companies cannot allow China to breach internation laws and human rights as it wishes. Th e long term ramifications to this are unthinkable, as China gains further strength. If Google wants to take its place as the greatest company in the world, surpassing Microsoft, it should stake its moral ground now.

A Job for Predictive Analytics; increased cost for Prescription Drugs

Jeffrey Newman — Wed, 13 Jan 2010 01:36:43 +0000

The steady and significant increases in the price of medicines, which has no indication of slowing, cries out for an extrapolation of costs going forward. Statheads come forth. Between October 2008 and September 2009, the brand-named medications most commonly used by Medicare beneficiaries rose an average of 9.3 percent. Interestingly, this was a period when there was no inflation and when consumer prices actually dropped. Those drugs that spiked their prices the most included Seroquel (to treat psychosis) 16.1%; Aricept(for dimentia), 17.2% increase; Ambien (for insomnia) 18.8% increase and Flomax (prostate) a 19.7% increase. If you analyze the drug industry answers to why such stark (and obscene) increases in prices of these and other medications, they say that it costs more to make drugs now. 18% per year more? BS! Of course the startling increases affect those on fixed incomes and those who are disabled and those who won’t get cost of living increases. Here’s where the statheads come in. The use of predictive analysis tools that have been developed including machine learning would allow us to predict with a high degree of probability the cost of most drugs separately as they will increase for the next five, ten and twenty years based on historical figures. This information, it seems to me, would be quite valuable in the hands of economists who can also predict the percentage of the population who will and will not be able to afford their medications. There are public databases of information with all information needed to determine the trend. Anyone game?

SOCIAL MEDIA TO PREVENT MURDERS AND SUICIDES?

Jeffrey Newman — Mon, 14 Dec 2009 20:49:57 +0000

This tidbit of creative thinking was found on the site of Jonathan Nguyen, who works in PR andcrises management. He saw a news story about George Sodini, who murdered three and wounded nine others in a Pennsylvania gym before killing himself. Sodini had talked about his plans on his blog, as have many others. Now, there are social media monitoring tools such as Radian6 which can read and listen to conversations on the web, including the blogosphere via an heuristic analysis engine. It is used sometimes to protect brands from disparagement and to guage sentiment on products. The questions raised is whether it could be modified to watch for potential homicidal or suicidal acts? Outstanding lateral thinking Jonathan! Present uses of social media are rudimentary and limited.

CAN STATISTICAL MODELS PREDICT THE GLOBAL FUTURE?

Jeffrey Newman — Fri, 13 Nov 2009 17:43:35 +0000

A 2007 article in Good Magazine summarized the work of Bruce Bueno de Mesquita who claims to predict mathematically, “virtually any international conflict, provided the basic inputs are accurate.” Bueno de Mesquita can also be seen on You Tube giving lectures. His work, which some dismiss too readily, raises important issues concerning the use of statistics and game theory to reliably envision the probable outcome of major world events. Who is this guy? He’s an NYU Professor of Political Science who has also studied statistics and has created a model simulation program. In the program, he takes a particular world flashpoint and plots out the probable way things will unfold based on what is known of the nation-state leaders and their actions assuming they work from a point of self interest. He acknowledges the existence of Black Swans, the events we can never anticipate but says their existence does not do away with his ability to determine a probable scenario and outcome. Allan Stam, Professor of Political Science at Penn State, examined the validity of Bueno de Mesquita’s work in the October 2000 British Journal of Political Science. He supported Bueno de Mesquita’s work, within certain limitations. A year before, Stephen Walt on the faculty at Harvard, attacked the assumptions of Bueno de Mesquita’s approach using “rational choice” models. (players promoting their own interests which Walt says is not always the case, especially when emotions or mental illness come to play). Bueno de Mesquita is asserting that consistent and repeatable findings that can reliably determine the outcome of major international conflicts. What has yet to be considered is what happens when other people use the same model? Can they achieve the same results consistently, using the same model and methods? Given that Bueno De Mesquita is claiming to use the scientific model, this, measure is the essential measure of his work. Looking more broadly, human behavior is obviously difficult to predict. However, decisions made by world leaders in crises do seem to have their own internal logic and statistical modeling is advancing swiftly in sophistication. Here where I think all of this fits. If Bueno de Mesquita and his model could reliably predict the outcome of international crisis points, he would be one of the wealthiest and most influential persons on the planet, which he isn’t. However, we now have an enhanced and growing ability to amass volumes of information quickly and to access meaningful data previously kept secret. This, combined with an increasing knowledge about the dynamics of world affairs and the internal workings of nations may allow us to reach a point where these kinds of model simulations will set a reliable base point for projecting of world trends. (As an aside, I think Bueno de Mesquita would make a great character for a dystopian sci fi novel placing him into the time when such models work and showing what horror it would wreak on his life and the lives of those around him. Sometimes, it doesn’t pay to know too much). What do you think?

STAR TREK THE FUTURE TELESCOPES TO THE PAST

Jeffrey Newman — Fri, 05 Jun 2009 15:30:00 +0000

Meet The New Spock - Same As The Old Spock?

This is my first blog about the future. Yesterday, I went to see Star Trek and although my expectations were tempered from prior Star Trek films, I came away thinking that this one had a lot to it and a lot that is relevant to this particular point in human history, where all that is past and all that will be future are on a direct collision course. Without giving away the plot, there is a point in the film where the elder Spock travels back in time and meets himself, the young Spock. It seems to me that given the state of the planet, we are all facing our younger selves and must ask the question, what can I do to help protect our fragile future.

I am now obsessed with all things future. I am reading about predictive analysis and machine learning and the immense potential that these endeavors have on our ability to reliably model global trends and changes. This is an arena of hope which may help us work through the more difficult problems of environmental damage and alternative energy. I recommend the book Super Crunchers by Ian Ayres which explains the impact that analytics is having on our society.
The question remains however as to whether we have enough time to catch up with the dilemmas we have created.

I have read Nassim Taleb’s The Black Swan which explains why over-reliance on statistical models can lead to destruction. However, enhanced data analysis may enable us to make changes more efficiently and using fewer resouces. This is especially the case in medicine and Ayres book explores the use of statistical analysis to seek out trends in human illness (AIDS/HIV).

See you in the future