Skip to main content
DSKim's user avatar
DSKim's user avatar
DSKim's user avatar
DSKim
  • Member for 11 years
  • Last seen more than a month ago
37 votes
6 answers
47k views

Why should we shuffle data while training a neural network?

30 votes
1 answer
59k views

Hessian of logistic function

9 votes
1 answer
7k views

full batch vs online learning vs mini batch

8 votes
2 answers
2k views

normalization in max-sum algorithm (loopy belief propagation)

8 votes
2 answers
1k views

How are the convergence conditions / KKT conditions for the soft-margin SVM derived

7 votes
2 answers
3k views

Solving constrained optimization problem: projected gradient vs. dual?

6 votes
2 answers
2k views

Maximum Entropy and Multinomial Logistic Function

3 votes
1 answer
570 views

why does drop-out increase training time?

2 votes
1 answer
920 views

why contrastive loss for siamese network

2 votes
1 answer
784 views

definition of deep belief network

2 votes
1 answer
406 views

train / validation / test split

2 votes
1 answer
50 views

Benefits of warmup in optimization

1 vote
2 answers
3k views

Shape of the hidden state of LSTM in Keras [closed]