pir's user avatar
pir's user avatar
pir's user avatar
pir
  • Member for 9 years, 1 month
  • Last seen more than a month ago
49 votes
7 answers
47k views

Importance of local response normalization in CNN

44 votes
3 answers
43k views

Are pooling layers added before or after dropout layers?

44 votes
2 answers
18k views

Difference between LOESS and LOWESS

26 votes
1 answer
31k views

What is a feasible sequence length for an RNN to model?

23 votes
3 answers
16k views

Importance of the bias node in neural networks

21 votes
1 answer
36k views

Preventing overfitting of LSTM on small dataset

14 votes
2 answers
4k views

Capturing initial patterns when using truncated backpropagation through time (RNN/LSTM)

14 votes
3 answers
11k views

Number of bins when computing mutual information

13 votes
2 answers
618 views

Different AIC definitions

12 votes
2 answers
7k views

Purpose of L2 normalization for triplet network

12 votes
2 answers
20k views

Handling unknown words in language modeling tasks using LSTM

12 votes
1 answer
2k views

Why are activation functions needed in neural networks? [duplicate]

12 votes
1 answer
3k views

Gradient clipping when training deep neural networks

11 votes
1 answer
16k views

Oversampling with categorical variables

10 votes
2 answers
9k views

L2-norms of gradients increasing during training of deep neural network

10 votes
1 answer
4k views

When to use LDA over GMM for clustering?

9 votes
2 answers
5k views

Receptive field of neurons in LeNet

8 votes
0 answers
979 views

Feature scaling for non-negative sparse data

8 votes
4 answers
592 views

Clusterings that can be caused by K-means

8 votes
2 answers
15k views

How do I force the L-BFGS-B to not stop early? Projected gradient is zero

8 votes
3 answers
759 views

Goodness of fit test on sparse contigency tables with high dimensionality

7 votes
1 answer
3k views

Limitations to generalized additive model (GAM)

6 votes
1 answer
1k views

Visualizing relationship between independent variable and binary response

6 votes
1 answer
903 views

How does backpropagation learn convolution filters?

6 votes
0 answers
291 views

When to choose GAM over LOESS? [duplicate]

5 votes
3 answers
194 views

Controlling for non-linear variable in non-linear modeling of response

5 votes
3 answers
2k views

Validation loss going down, but validation accuracy worsening

5 votes
3 answers
1k views

Removing attributes with few observations in R

5 votes
1 answer
579 views

Placement of earlier features in more complex features in CNN

5 votes
1 answer
2k views

Convolutional neural network - Using absolute of tanh on convolution output