In October, Nasdaq, the stock exchange, introduced a tool that USES deep learning in artificial intelligence to flag suspicious trades. The technology should make it easier to spot scammers in billions of transactions a year.
Nasdaq has adopted new software that, along with traditional methods, looks for stock market crimes such as insider trading. Michael O ‘rourke, nasdaq’s director of machine learning, says deep learning is important because it is “good at finding things that are hard to describe”.
To enable the software to enhance deep learning, nasdaq provides it with exchange order and trading data, as well as non-public information. After a year of testing, the team that created the technology decided it was reliable enough for widespread use.
One potential risk of using artificial intelligence to spot a scam is that it can generate so many false positives that it overwhelms the people in charge of vetting the deals. “You can’t miss it, but you don’t want to kill it either,” o ‘rourke said. However, he added that it was too early to calculate a meaningful margin of error, and that he found a certain number of miscalculations acceptable.
Ultimately, the onus is on federal regulators to track down offenders, a challenge for the Securities and Exchange Commission, which has limited resources. Still, the nasdaq team says the new software helps their work by providing better, more timely information.
RC Blogs 