User Firebug

73 votes

Accepted

40,000 neuroscience papers might be wrong

answered Jul 25, 2016 at 17:16

61 votes

Accepted

Practical hyperparameter optimization: Random vs. grid search

answered Apr 26, 2016 at 15:19

48 votes

Accepted

Calculating PCA variance explained

answered Jan 4, 2017 at 19:49

42 votes

Accepted

Does random forest need input variables to be scaled or centered?

answered Jan 12, 2017 at 2:22

33 votes

Accepted

hinge loss vs logistic loss advantages and disadvantages/limitations

answered Jul 20, 2016 at 20:55

29 votes

Accepted

What's considered a good log loss?

answered Apr 26, 2017 at 18:08

29 votes

Accepted

How to determine the accuracy of regression? Which measure should be used?

answered Sep 13, 2018 at 20:12

25 votes

Accepted

How does `predict.randomForest` estimate class probabilities?

answered Jul 28, 2016 at 20:58

17 votes

Random forest regression not predicting higher than training data

answered Sep 15, 2016 at 18:28

17 votes

What important ideas came since Nelder and McCullagh's book Generalized Linear Models (a 40 year old book)?

answered May 1, 2022 at 13:10

16 votes

Accepted

Is Gradient Descent possible for kernelized SVMs (if so, why do people use Quadratic Programming)?

answered Aug 18, 2016 at 21:17

16 votes

Why isn't variance defined as the difference between every value following each other?

answered Jul 26, 2016 at 17:50

15 votes

Accepted

How to obtain optimal hyperparameters after nested cross validation?

answered Jan 4, 2017 at 23:31

14 votes

Accepted

T-test paradox: can adding a single point very far from the null value change the outcome from significant to nonsignificant?

answered Apr 25, 2017 at 22:31

14 votes

Accepted

Is there any classification algorithm that doesn't give probability?

answered Feb 21, 2020 at 14:09

14 votes

Why does regularization wreck orthogonality of predictions and residuals in linear regression?

answered Oct 30, 2020 at 0:47

13 votes

How should I transform non-negative data including zeros?

answered Jul 5, 2016 at 13:57

13 votes

Accepted

Modern machine learning and the bias-variance trade-off

answered Jun 23, 2021 at 12:07

12 votes

When would you use PCA rather than LDA in classification?

answered Sep 20, 2016 at 13:54

11 votes

Accepted

Supervised learning with uncertain data?

answered Jan 7, 2017 at 3:55

11 votes

Accepted

Meaning of Epsilon in SVM regression

answered Jan 31, 2017 at 12:21

11 votes

What does the "machine" in "support vector machine" and "restricted Boltzmann machine" mean?

answered Feb 23, 2017 at 19:42

11 votes

Accuracy vs. area under the ROC curve

answered Jul 23, 2016 at 2:45

11 votes

How can I better understand this covariance equation?

answered Oct 14, 2021 at 19:55

10 votes

Classification with ordered classes?

answered Jul 4, 2016 at 14:08

10 votes

Accepted

What does it mean if I have a high F-stat but low $R^2$?

answered Oct 8, 2020 at 18:04

9 votes

Misconception about ReLu

answered Jul 26, 2021 at 18:56

9 votes

Accepted

Negative R squared but with close predictions

answered Feb 10 at 10:39

9 votes

Accepted

$\text{Var}(y)$ in linear regression

answered Nov 4, 2020 at 22:22

8 votes

How do I determine whether two correlations are significantly different?

answered May 10, 2017 at 20:29

Stack Exchange Network

388 Answers

40,000 neuroscience papers might be wrong

Practical hyperparameter optimization: Random vs. grid search

Calculating PCA variance explained

Does random forest need input variables to be scaled or centered?

hinge loss vs logistic loss advantages and disadvantages/limitations

What's considered a good log loss?

How to determine the accuracy of regression? Which measure should be used?

How does `predict.randomForest` estimate class probabilities?

Random forest regression not predicting higher than training data

What important ideas came since Nelder and McCullagh's book Generalized Linear Models (a 40 year old book)?

Is Gradient Descent possible for kernelized SVMs (if so, why do people use Quadratic Programming)?

Why isn't variance defined as the difference between every value following each other?

How to obtain optimal hyperparameters after nested cross validation?

T-test paradox: can adding a single point very far from the null value change the outcome from significant to nonsignificant?

Is there any classification algorithm that doesn't give probability?

Why does regularization wreck orthogonality of predictions and residuals in linear regression?

How should I transform non-negative data including zeros?

Modern machine learning and the bias-variance trade-off

When would you use PCA rather than LDA in classification?

Supervised learning with uncertain data?

Meaning of Epsilon in SVM regression

What does the "machine" in "support vector machine" and "restricted Boltzmann machine" mean?

Accuracy vs. area under the ROC curve

How can I better understand this covariance equation?

Classification with ordered classes?

What does it mean if I have a high F-stat but low $R^2$?

Misconception about ReLu

Negative R squared but with close predictions

$\text{Var}(y)$ in linear regression

How do I determine whether two correlations are significantly different?