mon's user avatar
mon's user avatar
mon's user avatar
mon
  • Member for 5 years
  • Last seen this week
14 votes
4 answers
68k views

Meaning of 'hue" in seaborn barplot

7 votes
1 answer
651 views

Why Batch Normalization is undesirable?

3 votes
2 answers
74 views

Result of uniform weight initialization in all neurons

3 votes
2 answers
15k views

Supported GPU for Pytorch

3 votes
2 answers
2k views

Gradient descent implementation of logistic regression

2 votes
1 answer
2k views

Batch normalization for image CNN - Why not use the mean of the entire batch?

2 votes
1 answer
6k views

How to make scipy.optimize.basinhopping find the global optimal point

2 votes
1 answer
3k views

Loss function for ReLu, ELU, SELU

1 vote
0 answers
5k views

How to understand the return values of scipy.interpolate.splrep

1 vote
1 answer
239 views

Can Batch Normalization replace tanh in RNN?

1 vote
0 answers
145 views

DCGAN - advise on why the training is not working

1 vote
2 answers
693 views

How to decide the number of primary components for PCA

1 vote
0 answers
20 views

Implementing eigen decomposition

1 vote
1 answer
35 views

Annotated Transformer - Why x + DropOut(Sublayer(LayerNorm(x)))?

0 votes
1 answer
31 views

Why the standard deviation of the BERT weight initialization is 0.02 by default

0 votes
0 answers
42 views

Is covariance matrix required to compute SVD?

0 votes
1 answer
148 views

What may cause the CNN layer weight regularizer to reduce the model accuracy

0 votes
1 answer
86 views

What is RFormula in simple English?

0 votes
1 answer
387 views

Transformers - Why Self Attention calculate dot product of q and k from of same word?

0 votes
1 answer
1k views

SparseCategoricalcrossEntropy(from_logits=True) internally apply softmax?

0 votes
1 answer
777 views

Clarification on tf.keras.metrics.AUC

0 votes
2 answers
5k views

seaborn regplot does not show the line

0 votes
1 answer
315 views

matplotlib subplots_adjust - meaning of parameters

0 votes
1 answer
16 views

Why no scale parameter for skip connection addition?