Deep Learning for Computer Vision and Natural Language Processing

• Schedule
• Reading
• Resource
• Projects
• Prog. Problem

Problem

In class programming problem: classifying 14 x 14 MNIST images

• MNIST small (14x14)
• How to prepare MNIST small and test it with a MLP experiment

After class improvement

• Chad DeChant: 0.833% with dropout. See Chad's code
• Chad DeChant: 0.95% testing error with RMSprop (introduction)
  • no pooling in conv layers
  • smaller conv filters: 3x3
  • rectified linear units
  See Chad's code.
• Christopher Cleveland and Zheng Shou: 1.3%

Liangliang Cao and James Fan, Updated 02/21/2015