It is a matter of time. ‘As the machines take over on world stock markets

Ke Jie was once the best complex in the world game player will ever invented, an ancient Chinese board game called Go. But in 2017, Ke was beaten by a computer program, which was taught how to play the game Ke most of the times he had spent his life mastering. The robot’s victory marked a turning point for Artificial Intelligence (AI) and machine learning technology-one to perform subset of artificial intelligence, with a computer learning without a task explicitly be programmed to complete, rather than learning and improve the experience. Now the technology is applied in the market sector, where algorithms are used to teach self-driving cars, such as busy streets to explore the city, to health care, in which robots learn to diagnose and treat patients. And in finance, increasingly, of which stocks to buy and sell these decisions technologies. Marcos López de Prado was in the front line of machine learning innovation in finance. The Spaniard has been based in New York before leaving earlier this year to start the first machine learning upside AQR, one of the largest investment management company in the world his own company, such as learning Know -Car and algorithms for Wall Street sold. The financial pioneer literally wrote the book on the use of the machine to learn how to invest (its progress 400 pages of textbooks in financial machine learning is included in the program of a number of graduate school courses), and he was the 2019 ” Quant of the Year “by the Journal of portfolio management. TIME sat down with López de Prado when he recently visited Hong Kong. Here is what one of the quants best the world has to say about how robots take over the global financial markets, and his high hopes for the technology. What progress do you see in the field of machine learning? Machine learning has been very successful over the expectations of many people in various industrial applications. A few years ago, is part of all this progress, as contained machines, the best player on [board game] Go defeat. What has changed is that we who are strong enough to solve these problems for computers. And we have a lot of data that did not exist before and we know how this data in combination use with algorithms. Then, all of a sudden, machine learning algorithms have tasks to experts, where it could be five years ago only human experts. What are the challenges of using this technology for buying stock and selling? Finance is special because they do not have large amounts of data than the data used to make these algorithms to train for the things to recognize as faces or cars. And develop markets. Are you an investor if you want to extract money from the market, the market learns to avoid extracting profits next year. Continue reading: Artificial Intelligence is powerful-and misunderstood. Here’s how we protect workers see what the mistakes when applying human machine learning to its investment process? Oracle This approach uses machine learning as a black box, and is doomed to fail. What makes no sense is to make at this point, to make these techniques to use Blackbox forecasts. Unless you understand why the model works in the first place, you do not know when to turn it off. What is the right way to invest in learning to use the machine? Machine learning is to be used as a research tool, not as a forecasting tool. Should new theories are used to identify, and if you identify a new theory, take the car, you do not want the car. In my opinion this is the right approach in the financial sector. Everyone should use machine learning theories and testing to develop theories, but once the theory is revealed, you must run the machine, not the machine. As technology changed the Stock Market? Today, most of the operations are performed by algorithms. There are very, very few transactions that are carried out by people. People have been sold. Anyone with the effectiveness of these and machine learning in doubt should look at runtime. It will completely take over the machines to invest? There is no reason not to expect that in the end these machines will be able to solve many tasks that we can not solve, then it is a matter of time, is not a matter of thinking though. I sometimes take the cars investing. When it is anyone’s guess. Decisions will process it for information. If you have a car, the information can be processed faster and better and objectively, then you will succeed. Will the company still money to the location where the cars make the investment? I think ultimately will have a very positive effect on society. If the activities more efficiently allocated, companies to be more successful, to get the goods and companies that are not very likely to succeed, will not receive goods. Why is it important for the development of this technology? My hope is that when these technologies are used properly, they will make markets more efficient so that when people invest will invest the following scientific approaches, instead of making speculations or wild wild conjectures, and it is hoped that the company of which helps the that you want something as important as finance and investments justified by scientific evidence, not mere speculation.
Picture Copyright by Klaus Ohlenschläger-picture-alliance / dpa / AP Images