Artificial Intelligence via Deep Learning
Abstract: Artificial Intelligence (AI) remains to be one of the most fascinating topics across computer science, neural science, psychology, statistics, applied mathematics, optimization theory, and even philosophy. After a half century of up and down, with some excitements but mostly frustrations, we begin to see very significant progresses on speech recognition, image recognition, machine translation, gaming agents, autonomous driving etc.. In particular, by leveraging big data aided by massive parallel computation, deep learning has played a very central role in pushing the boundary of all these AI areas. I will introduce some latest technology achievements, and briefly discuss the computational and statistical factors behind this phenomenon success. More importantly, we will see these technologies have been shaping up a foundation for imaginative long-term innovation.
Time: Oct. 20, 2014, 2:30 pm
Venue: A601, NAOC