[ad_1]
This decade of data is what drives the university’s new experiment in artificial intelligence.
Dr. Finn and his team are a plexusA mathematical system that can learn skills from large amounts of data. A neural network can learn to identify a cat by pinpointing patterns in thousands of cat photos. It can learn to recognize spoken words by analyzing hundreds of old phone calls. Or, by studying how lecturers evaluate coding tests, they can learn to evaluate these tests on their own.
The Stanford system spent hours analyzing samples from old midterms, learning from ten-year probabilities. Then he was ready to learn more. Given just a handful of extra examples from the new exam presented this spring, he can quickly grasp the task at hand.
“He sees many kinds of problems,” said Mike Wu, another researcher working on the project. “Then he can adapt to problems he has never seen before.”
According to a study by Stanford researchers, this spring, the system provided 16,000 pieces of feedback, and students accepted 97.9 percent of the feedback. By comparison, students participated in feedback from human instructors 96.7 percent of the time.
Mr. Pham, an engineering student at Lund University in Sweden, was surprised that the technology was working so well. While the automated tool wasn’t able to evaluate one of its programs (probably because it wrote a code snippet unlike anything the AI had ever seen), both detected certain errors in its code, including those known in computer programming and math. a fence post fault and suggested ways to fix them. “You rarely get feedback so well thought out,” said Mr Pham.
Technology was effective because its role was so sharply defined. Mr. Pham wrote code for very specific purposes while taking the test, and there were many ways he and other students could go wrong.
But given the right data, neural networks can learn a range of tasks. It’s the same basic technology Identifies faces in photos you post to Facebook, recognizes commands you bark into your iPhone and translates from one language to another On services like Skype and Google Translate. The hope for the Stanford team and other researchers is that these techniques can automate training in many other ways.
[ad_2]
Source link