User Matthew Drury

19 votes

Accepted

What is use of Tweedie or poisson loss/objective function in XGboost and Deep learning models

answered Oct 19, 2020 at 18:45

4 votes

A question about conditional probability

answered Oct 17, 2017 at 5:26

69 votes

Accepted

Why use regularisation in polynomial regression instead of lowering the degree?

answered Jul 31, 2016 at 16:34

43 votes

Accepted

Understanding complete separation for logistic regression

answered Jul 21, 2016 at 4:37

3 votes

How to "squish" a range of numbers to the same spread between 0 and 1

answered Aug 6, 2021 at 1:42

6 votes

Accepted

GBM: impact of the loss function

answered Jun 5, 2016 at 23:03

2 votes

Accepted

Evaluating Logistic Regression

answered Aug 29, 2020 at 16:35

7 votes

Accepted

Visualizing the hinge loss and 0-1 loss

answered Aug 3, 2020 at 0:02

5 votes

Accepted

Understanding the weak learners in boosting

answered Jun 25, 2020 at 5:07

2 votes

Applying de-standardised ridge regression coefficients to new test data - how to best handle the mean of y_test?

answered May 8, 2020 at 16:42

3 votes

Can someone explain why neural networks are highly parameterized?

answered Apr 21, 2020 at 4:32

3 votes

Adjusting classification thresholds based on test set predictions

answered Mar 6, 2020 at 17:04

6 votes

Does XGBoost has anything to do with logistic regression?

answered Feb 27, 2020 at 15:30

4 votes

Accepted

Uniform random variable is greater by a constant from another uniform random variable

answered Feb 22, 2020 at 20:38

6 votes

Accepted

For simple linear regression, is $\beta_1$ linear in $y_i$?

answered Feb 21, 2020 at 4:02

3 votes

Is there a bootstrap 're-sampling the residuals' equivalent for GLM?

answered Feb 7, 2020 at 2:59

14 votes

Accepted

Do all bounded probability distributions have a definite mean?

answered Dec 26, 2019 at 17:30

3 votes

Accepted

Variable is not statistically significant in single variable linear regression but is significant in multiple regression

answered Dec 18, 2019 at 6:03

1 vote

Machine Learning - Bagging

answered Aug 10, 2019 at 4:27

1 vote

Is there a Status Machine Chart?

answered Aug 5, 2017 at 22:49

55 votes

Accepted

How to apply standardization/normalization to train- and testset if prediction is the goal?

answered Sep 30, 2015 at 16:16

2 votes

Model Decreasing in Accuracy With More Training Data

answered Apr 3, 2019 at 16:04

2 votes

disadvantages of svm

answered Apr 3, 2019 at 3:43

1 vote

Accepted

Should loss function be defined over output or parameters?

answered Mar 27, 2019 at 3:25

42 votes

Accepted

Is a model fitted to data or is data fitted to a model?

answered Mar 24, 2019 at 4:44

9 votes

Need help understanding what a natural log transformation is actually doing and why specific transformations are required for linear regression

answered Mar 22, 2019 at 23:52

19 votes

Accepted

Why does Central Limit Theorem break down in my simulation?

answered Mar 9, 2019 at 4:01

1 vote

Accepted

2-step linear regression

answered Mar 10, 2019 at 17:47

10 votes

Accepted

Does the universal approximation theorem for neural networks hold for any activation function?

answered Jan 30, 2018 at 5:21

5 votes

Accepted

Does retraining a model on all available data necessarily yield a better model?

answered Mar 4, 2019 at 23:53

Stack Exchange Network

432 Answers

What is use of Tweedie or poisson loss/objective function in XGboost and Deep learning models

A question about conditional probability

Why use regularisation in polynomial regression instead of lowering the degree?

Understanding complete separation for logistic regression

How to "squish" a range of numbers to the same spread between 0 and 1

GBM: impact of the loss function

Evaluating Logistic Regression

Visualizing the hinge loss and 0-1 loss

Understanding the weak learners in boosting

Applying de-standardised ridge regression coefficients to new test data - how to best handle the mean of y_test?

Can someone explain why neural networks are highly parameterized?

Adjusting classification thresholds based on test set predictions

Does XGBoost has anything to do with logistic regression?

Uniform random variable is greater by a constant from another uniform random variable

For simple linear regression, is $\beta_1$ linear in $y_i$?

Is there a bootstrap 're-sampling the residuals' equivalent for GLM?

Do all bounded probability distributions have a definite mean?

Variable is not statistically significant in single variable linear regression but is significant in multiple regression

Machine Learning - Bagging

Is there a Status Machine Chart?

How to apply standardization/normalization to train- and testset if prediction is the goal?

Model Decreasing in Accuracy With More Training Data

disadvantages of svm

Should loss function be defined over output or parameters?

Is a model fitted to data or is data fitted to a model?

Need help understanding what a natural log transformation is actually doing and why specific transformations are required for linear regression

Why does Central Limit Theorem break down in my simulation?

2-step linear regression

Does the universal approximation theorem for neural networks hold for any activation function?

Does retraining a model on all available data necessarily yield a better model?