Computers having Regrets?
Regret is an important part of adaptation, and most people learn not to repeat the same mistakes. But for computers, this kind of hindsight does not occur—the same computing task could fail many times without being adjusted. Now, Google has teamed up with researchers at Tel Aviv University to ask the question “could this hindsight be programmed into a computer to more accurately predict the future?”
Professor Yishay Mansour of Tel Aviv University's Blavatnik School of Computer Science is leading the project, which was launched this year at the International Conference on Learning Theory in Haifa, Israel. Google is funding the research team, which consists of computer scientists, economists, and game theorists, to help computers minimize “regret.”
"If the servers and routing systems of the Internet could see and evaluate all the relevant variables in advance, they could more efficiently prioritize server resource requests, load documents and route visitors to an Internet site, for instance," Mansour said. This kind of efficiency is attractive to companies like Google, which uses massive resources to keep its sites running. The research could particularly benefit its AdWords and Adsense advertising platforms by helping computers choose more effective displays.
Mansour defines computer “regret” as the discrepancy between the desired and the actual outcome of a task. He recently created an artificial intelligence algorithm to help minimize this discrepancy.
"We are able to change and influence the decision-making of computers in real-time,” Mansour said. “Compared to human beings, help systems can much more quickly process all the available information to estimate the future as events unfold — whether it's a bidding war on an online auction site, a sudden spike of traffic to a media website, or demand for an online product.”
If successful, the algorithm will be able to adapt as it is running. According to Mansour, this means that a task runs "almost as if you knew all the variables in advance.”
The 20-person team includes scientists from Tel Aviv University, Hebrew University, and Google. Yossi Matias, a Tel Aviv University faculty member, is also the head of Google Israel.
By combining academic prowess with industry experience, the team hopes to profit many different aspects of the Internet. "We are asking how we can give incentives to get bidders and buyers in the auction to behave intelligently, by understanding the dynamics of the auction process," said Mansour.
Professor Yishay Mansour of Tel Aviv University's Blavatnik School of Computer Science is leading the project, which was launched this year at the International Conference on Learning Theory in Haifa, Israel. Google is funding the research team, which consists of computer scientists, economists, and game theorists, to help computers minimize “regret.”
"If the servers and routing systems of the Internet could see and evaluate all the relevant variables in advance, they could more efficiently prioritize server resource requests, load documents and route visitors to an Internet site, for instance," Mansour said. This kind of efficiency is attractive to companies like Google, which uses massive resources to keep its sites running. The research could particularly benefit its AdWords and Adsense advertising platforms by helping computers choose more effective displays.
Mansour defines computer “regret” as the discrepancy between the desired and the actual outcome of a task. He recently created an artificial intelligence algorithm to help minimize this discrepancy.
"We are able to change and influence the decision-making of computers in real-time,” Mansour said. “Compared to human beings, help systems can much more quickly process all the available information to estimate the future as events unfold — whether it's a bidding war on an online auction site, a sudden spike of traffic to a media website, or demand for an online product.”
If successful, the algorithm will be able to adapt as it is running. According to Mansour, this means that a task runs "almost as if you knew all the variables in advance.”
The 20-person team includes scientists from Tel Aviv University, Hebrew University, and Google. Yossi Matias, a Tel Aviv University faculty member, is also the head of Google Israel.
By combining academic prowess with industry experience, the team hopes to profit many different aspects of the Internet. "We are asking how we can give incentives to get bidders and buyers in the auction to behave intelligently, by understanding the dynamics of the auction process," said Mansour.
Comments