Differences

This shows you the differences between two versions of the page.

Link to this comparison view

wide_and_deep_learning [2017/08/31 10:36]
127.0.0.1 external edit
wide_and_deep_learning [2018/04/15 22:35]
admin
Line 183: Line 183:
  
 https://​my.memo.ai/​external/​vJ-iBYK4TjS8aAZMj8LP Deep & Cross Network for Ad Click Predictions (Google 2017) https://​my.memo.ai/​external/​vJ-iBYK4TjS8aAZMj8LP Deep & Cross Network for Ad Click Predictions (Google 2017)
 +
 +https://​arxiv.org/​pdf/​1804.03782.pdf CoT: Cooperative Training for Generative Modeling
 +
 +We proposed Cooperative Training, a powerful unbiased, low-variance,​ computationally efficient
 +algorithm inspired by coordinate decent algorithms like GAN and Expectation Maximization. Models
 +trained via CoT shows promising results in many sequential data modeling tasks. ​ In this paper, we propose Cooperative Training (CoT), an efficient, low-variance,​ bias-free algorithm
 +for training likelihood-based models by directly optimizing a well-estimated Jensen-Shannon divergence.
 +CoT coordinately trains a generative module G, and an auxiliary predictive module M,
 +called mediator, for guiding G in a cooperative fashion
 +