A neural network learns when it should not be trusted

MIT researchers have developed a way for deep learning neural networks to rapidly estimate confidence levels in their output.

The advance could enhance safety and efficiency in AI-assisted decision making.

Credit - Image: iStock image edited by MIT News

News Source: https://news.mit.edu/2020/neural-network-uncertainty-1120

Playlist for Technology News Videos - https://www.youtube.com/playlist?list=PLK2ccNIJVPpB_XqWWq_oaZGIDzmKiSkYc Artificial Intelligence - https://www.youtube.com/playlist?list=PLK2ccNIJVPpBqKHdezLeaUzIp85DxFu7_ ... https://www.youtube.com/watch?v=ggQaARlLzEE