Learning the Power of Teamwork in a Netflix Race for $1 Million
A competition has installed from Netflix, than he has not offered to a prize $1 million to that he could meaningfully improve the relative system of film recommendation, concluded Sunday with two squares in a virtual and winning out of order heat to declare till september.
Jump to the following paragraph
Ingrandica this image
Ozier Muhammad/New York Times
Chris Volinsky, a scientist to AT& amp; The search of T, of left, is on a high square in a competition of Netflix. With he it is Robert Bell.
Relative
Subjects of periods: Netflix Inc.
NYT_VideoPlayerStart ({
playerType: & quot; article& quot; ,
videoId : & quot; 1194833560594& quot;
});
But the competition, that he has begun nell' October 2006, already has produced un' impressive inheritance. It has modeled the careers, placed eggs at least un' company start-up and inspired the banns. Moreover it has changed the popular wisdom approximately the best sense to develop the automated systems that more and more help people to operate the choices online approximately films, books, garments, restaurants, news and other goods and services. These so-called motors of recommendation are computing the models that foretell of what a person could enjoy based on the signature statistics those person& amp; #8217; declared preferences s, chosen last and similar models of consumption operated from many others & amp; #151; entire it has rendered possible from the facility of the collection of data and dell' pursuit on the Web.& amp; #8220; The rewarded competition of Netflix will be examined for the years from the people that study like making the modellistica preventive, & amp; #8221; Said Chris Volinsky, a scientist to AT& amp; #38; Search of T and a head of one of the two squares high-aligned in the competition. The greatest learned lesson, second the members of the two advanced squares, was the collaboration power. It was not a single understanding, procedure or the concept that have allowed that both the squares surpassed l' Netflix objective, l' locativa film company, has regulated nearly three years ago: for in order at least improving the film recommendations made from the relative inner software from 10 for hundreds, like measured close foretold against the real appraisals of a-through-five-star from the customers. Instead, they say, the formula for succeeding era to re-unite people with the complementary abilities and to join the different methods of problem solving. This has become more and more appearing while the competition has evolven itself. Mr. Volinsky& amp; #8217; square of s, BellKor& amp; #8217; the pragmatic chaos of s, was the probable winner from long time and the first one in order to surpass the transenna of 10 for hundreds. It is really a collection of the seven-person of other squares and the relative members are statistical, expert in learning of machine and engineers in computer science from the United States, dall' Austria, from Canada and the Israel.When BellKor& amp; #8217; the s it has announced the past month that had passed the threshold of 10 for hundreds, it has provoked a race from 30 days, second the competition rules, so that other squares the tests at best. That one has lead to an other round one of square-fusion from BellKor& amp; #8217; the main rivals of s, that they have mounted a total consortium of approximately 30 members, with justice called the Ensemble.Submissions they have come fast and furious in last weeks from BellKor& amp; #8217; s and l' together. As soon as tiny before the expiration Sunday, the Ensemble& amp of competition; #8217; l' last entrance of s has bordered in front of BellKor& amp; #8217; s on the & amp public of the edge of the fotoricettore head; #151; from a-cent of a percentage point.& amp; #8220; The competition was nearly a race in order to agglomerate altretante squares like possible, & amp; #8221; Said David Weiss, a candidate of Ph.D in computer science all' University of the Pennsylvania and to a member dell' together. & amp; #8220; The surprise was that the collaboration method thus works well, what tries all the procedures, codifying she in on and putting far away have exceeded them with ours expectations.& amp; #8221; The changing contenders, seem, with the competition. When the competition of Netflix has begun, the Mr. Weiss was one of three old ones all' University of Princeton, comprised David Lin and Lester Mackey, that they have composed a called square Dinosaur Planet. The Mr. Lin, a greater one for the mathematics, has continued to become a trader of the bys-product on Wall Street. but the Mr. Mackey is a candidate of Ph.D to the statistical laboratory of artificial intelligence all' University of California, Berkeley. & amp; #8220; My interests now have been influenced working the rewarded competition of Netflix, & amp; #8221; it has said. The systems of recommendation of the software, the Mr. Mackey has said, more and more they will be transformed in tools common in order to help people to find the useful information and the products in means dell' outbreak of information and the offers that compete for their attention on the fotoricettore. & amp; #8220; Many these techniques will be propagheranno through the Internet, & amp; #8221; it has foretold. That is sure the hope of Domonkos Tikk, a computer science Hungarian and a member dell' together. The colleagues of the Mr. Tikk, 39 and three more young people have begun to little work to the competition time after that it began and in 2007 teamed in on with the group of the Princeton. & amp; #8220; When we have entered in the competition of Netflix, we have not had experience in the filtration of collaboration, & amp; #8221; The Mr. Tikk has said. However based on what they have learned, the Mr. Tikk and its connects has founded a departure, gravity, that it is putting to point the systems of recommendation for the customers trades them, comprised the Web site of electronic commerce and un' European company of the cellular one. Although the square dell' entirety touches with the elbow in front of BellKor& amp; #8217; s on the public edge of the head, he is not necessarily the winner. BellKor& amp; #8217; the s, according to the Mr. Volinsky, rests in the first place and Netflix it is put if it in contact with Sunday in order to say thus. And in a tribune online, an other member of the BellKor& amp; #8217; the square of s, Yehuda Koren, a investigator for Yahoo nell' Israel, has said that its square has had & amp; #8220; a better appraisal of the test that l' entirety, & amp; #8221; in spite of what the rival square has introduced for the edge of the head. Thus it is BellKor& amp; #8217; s the winner? Sure not still, second a spokesman of Netflix, Steve Swasey. & amp; #8220; Not there is winning, & amp; #8221; it has said. A winner, the Mr. Swasey has said, will probably not be announced until a moment within to september to an event accommodated from Reed Hastings, Netflix& amp; #8217; main executive of s. L' locativa film company is not outside holding for public relation maximum effect, the Mr. Swasey has said, but because the winner has still not been determined. The edge of the fotoricettore head, has explained, is based on what the squares introduces. After, Netflix& amp; #8217; the inside and investigators s expert in the external part must convalidare the teams& amp; #8217; presentations, poring over the introduced code, documents of design and other materials. & amp; #8220; This is roba really complex, & amp; #8221; The Mr. Swasey has said. In Hungary, the Mr. Tikk has not played optimistic. & amp; #8220; We didn& amp; #8217; the t the notification from Netflix obtains all, & amp; #8221; it has said in un' interview of I telephone. & amp; #8220; Thus task that the probabilities that we have gained or much light one. It was a try.& amp pleasant; #8221;