Multimodel DL to detect depression and intro to algorithmic thinking


Jul 21, 07:00 PM PDT
  • Virtual SF Bay ACM
  • 57 RSVPs
Description
This seminar is hosted by SF Bay ACM Chapter

An interdisciplinary framework is introduced to leverage both psychiatric clinical best practices as well as AI natural language processing and facial emotion recognition technologies. During COVID-19, the depression rate increased dramatically. Young adults are most vulnerable to the mental health effects of the pandemic. Sentiment analysis technologies and a uniquely designed depression Lexicon for young adults are applied to user communications on Twitter. The result is further integrated with emotion cues from Convolutional Neural Network. The experiment results are consistent with an average accuracy of 88.31%. This approach can reach 300+ million daily active Twitter users to promote early depression detection.

Algorithm is about working smart to avoid unnecessary work. It identifies the most efficient steps to solve a seemingly complex problem without detouring. This talk will use a unique and highly visual approach to introduce some fundamental algorithms used in computational competition and job interviews: Greedy, Dynamic Programming, Prim, Kruskal, Dijkstra, BFS, DFS, etc. Audience will be inspired by the Pygames and the algorithm inventors’ fun stories. You can be the next inventor!


Contact Organizer