Ben Reiniger
  • Member for 3 years, 6 months
  • Last seen this week
18 answers
142 votes
164k views
113 bookmarks
How do you visualize neural network architectures?
5 answers
54 votes
23k views
44 bookmarks
In softmax classifier, why use exp function to do normalization?
6 answers
53 votes
15k views
37 bookmarks
Should I go for a 'balanced' dataset or a 'representative' dataset?
4 answers
43 votes
19k views
18 bookmarks
Is it always better to use the whole dataset to train the final model?
4 answers
15 votes
34k views
12 bookmarks
L2 loss vs. mean squared loss
3 answers
14 votes
12k views
5 bookmarks
L1 & L2 Regularization in Light GBM
2 answers
11 votes
3k views
3 bookmarks
What is the proper way to use early stopping with cross-validation?
2 answers
10 votes
2k views
5 bookmarks
What is the most efficient method for hyperparameter optimization in scikit-learn?
2 answers
8 votes
103 views
1 bookmarks
Model for Differing Number of Rows per Observation
2 answers
8 votes
190 views
3 bookmarks
Time-series prediction: Model & data assumptions in AI/ML models vs conventional models
2 answers
8 votes
86 views
3 bookmarks
Which classification algorithms are negatively affected by class imbalances?
2 answers
8 votes
1k views
1 bookmarks
Transformer model: Why are word embeddings scaled before adding positional encodings?
2 answers
7 votes
266 views
1 bookmarks
Dropping features after final evaluation on test data
4 answers
7 votes
2k views
2 bookmarks
Can the vanishing gradient problem be solved by multiplying the input of tanh with a coefficient?
1 answers
6 votes
463 views
1 bookmarks
Strategies for handling unlabeled data which is slightly different from the labeled data
1 answers
6 votes
730 views
2 bookmarks
Understanding LSTM structure
1 answers
6 votes
432 views
1 bookmarks
Reticulate vs Python
1 answers
6 votes
1k views
2 bookmarks
Smart data split (train/eval) for Object Detection
2 answers
6 votes
200 views
1 bookmarks
ground truth fit is worse than cross validated fit on noisy data?
2 answers
5 votes
101 views
2 bookmarks
How to use "tree boosting" with a data-driven loss function
1 answers
5 votes
964 views
1 bookmarks
Multiclass classification on imbalanced dataset : Accuracy or micro F1 or macro F1
2 answers
5 votes
110 views
2 bookmarks
Does a precision score increasing with a higher number of folds mean the model will improve with more data?
3 answers
5 votes
899 views
5 bookmarks
Quasi-categorical variables - any ideas?
1 answers
5 votes
1k views
2 bookmarks
My CNN produces volatile validation_loss and does not converge in python(Keras)
1 answers
5 votes
159 views
1 bookmarks
Feature scaling worsens performance?
1 answers
4 votes
1k views
1 bookmarks
AttributeError: type object 'Lambda' has no attribute 'shape'
3 answers
4 votes
429 views
1 bookmarks
Train classifier on balanced dataset and apply on imbalanced dataset?
1 answers
4 votes
254 views
1 bookmarks
Hyperopt vs Default Values
2 answers
4 votes
2k views
4 bookmarks
Why does feature scaling improve the convergence speed for gradient descent?
1 answers
4 votes
4k views
1 bookmarks
svm.LinearSVC: larger max_iter number doesn't always increase the accuracy/precision/recall