user avatar
user avatar
user avatar
Eric Auld
  • Member for 4 years, 6 months
  • Last seen more than a week ago
6 votes
2 answers
955 views

Meaning of "identically distributed" when there's only one variable

6 votes
1 answer
627 views

Ridge regression: penalizing weights corresponding to larger-scale features

3 votes
1 answer
3k views

Weight scale inference when using dropout

3 votes
1 answer
13 views

Inputs not proxying for variance, but nevertheless well-correlated

2 votes
2 answers
145 views

Posterior of one observation transform into posterior of several observations

2 votes
1 answer
45 views

Modeling non-acyclic reality with acyclic graphs

2 votes
1 answer
1k views

Understanding dropout method: one mask per batch, or more?

2 votes
1 answer
790 views

Fast dropout: How to compute the mean and variance of the approximating Gaussian?

1 vote
1 answer
108 views

Indexing the formula for the ith layer of a linear deep neural network

1 vote
0 answers
776 views

Hinton claims SGD with batch norm can help: How?

1 vote
0 answers
144 views

Explaining batch normalization in terms of Gaussian processes

1 vote
1 answer
101 views

Bayesian model selection: picking the MAP model by integrating

1 vote
1 answer
318 views

Why is dropout causing my network to overfit so badly? [closed]

0 votes
0 answers
79 views

Joint distribution of estimator and true parameter