User einar

23 votes

Accepted

Why does finding small effects in large studies indicate publication bias?

answered Feb 20, 2019 at 12:36

22 votes

Accepted

Are Prophet's "uncertainty intervals" confidence intervals or prediction intervals?

answered Jun 27, 2023 at 11:06

22 votes

Accepted

How is empirical Bayes valid?

answered Aug 8, 2017 at 13:34

20 votes

Incorrectly Using the Word "Causal" to Describe a Regression Model?

answered May 9, 2023 at 8:49

18 votes

Why should boostrap sample size equal the original sample size?

answered Apr 25, 2017 at 12:05

14 votes

Accepted

Why do we say Bayesian statistics is suited for probability of one-time events?

answered Nov 13, 2020 at 11:47

13 votes

Accepted

Is it possible to decompose fitted residuals into bias and variance, after fitting a linear model?

answered Oct 26, 2017 at 10:35

11 votes

Accepted

Pearl, Causal Inference in Statistics Q3.5.1 (Backdoor criterion)

answered Jul 17, 2022 at 21:24

10 votes

Why is the posterior distribution the same as likelihood function when uniform prior distribution is used in Bayesian Analysis?

answered Aug 17, 2020 at 8:46

10 votes

A term for "number of columns" of a matrix

answered Nov 17, 2017 at 16:12

9 votes

Accepted

What's the point of reporting bootstrap bias?

answered Mar 6, 2018 at 17:03

8 votes

Accepted

Is it a confounder on not?

answered Oct 28, 2022 at 10:06

8 votes

Manually calculating logistic regression coefficient

answered Jun 23, 2016 at 12:58

7 votes

Accepted

Over-represented values in FDR-adjusted p-values

answered Feb 25, 2017 at 13:36

7 votes

Pros and cons of bootstrapping

answered May 20, 2017 at 15:00

6 votes

Statistical approach to compare the calibration between models

answered Dec 17, 2017 at 21:54

6 votes

Do I need a global test set when using k-fold cross-validation and a small n (n<<p)?

answered Feb 14, 2017 at 15:09

6 votes

In layman's terms, why is Naive Bayes the dominant algorithm used for text-classification?

answered Apr 16, 2017 at 10:31

6 votes

Why is lasso in matlab much slower than glmnet in R (10 min versus ~1 s)?

answered Nov 29, 2016 at 11:55

6 votes

Accepted

Matt's trick (reparametrization) makes my models slower, not faster

answered Jun 22, 2020 at 9:30

6 votes

Accepted

Why would Lasso recommend a lambda of 0?

answered Mar 5, 2021 at 7:33

5 votes

Accepted

For any distributions, are mean of the sum of the same distributions equals the sum of the means of the distributions?

answered Aug 1, 2019 at 7:25

5 votes

Bootstrap estimate for standard error in linear regression

answered Apr 22, 2017 at 11:37

5 votes

When is the bootstrap estimate of bias valid?

answered Oct 26, 2017 at 11:53

4 votes

How to make sense of this PCA plot with logistic regression decision boundary (breast cancer data)?

answered May 2, 2017 at 8:47

4 votes

How to explain low performance of naive Bayes on a dataset

answered Aug 21, 2017 at 10:39

4 votes

disadvantages variational inference

answered Oct 18, 2017 at 9:33

4 votes

Accepted

The results of CV on Ridge are different than the results of RidgeCV

answered Feb 20, 2017 at 11:22

4 votes

Does modeling with Random Forests require cross-validation?

answered Jul 21, 2015 at 10:49

4 votes

Accepted

Is one-hot encoding required for categorical variables in R (logistic regression)?

answered Feb 27, 2022 at 21:11

Stack Exchange Network

59 Answers

Why does finding small effects in large studies indicate publication bias?

Are Prophet's "uncertainty intervals" confidence intervals or prediction intervals?

How is empirical Bayes valid?

Incorrectly Using the Word "Causal" to Describe a Regression Model?

Why should boostrap sample size equal the original sample size?

Why do we say Bayesian statistics is suited for probability of one-time events?

Is it possible to decompose fitted residuals into bias and variance, after fitting a linear model?

Pearl, Causal Inference in Statistics Q3.5.1 (Backdoor criterion)

Why is the posterior distribution the same as likelihood function when uniform prior distribution is used in Bayesian Analysis?

A term for "number of columns" of a matrix

What's the point of reporting bootstrap bias?

Is it a confounder on not?

Manually calculating logistic regression coefficient

Over-represented values in FDR-adjusted p-values

Pros and cons of bootstrapping

Statistical approach to compare the calibration between models

Do I need a global test set when using k-fold cross-validation and a small n (n<<p)?

In layman's terms, why is Naive Bayes the dominant algorithm used for text-classification?

Why is lasso in matlab much slower than glmnet in R (10 min versus ~1 s)?

Matt's trick (reparametrization) makes my models slower, not faster

Why would Lasso recommend a lambda of 0?

For any distributions, are mean of the sum of the same distributions equals the sum of the means of the distributions?

Bootstrap estimate for standard error in linear regression

When is the bootstrap estimate of bias valid?

How to make sense of this PCA plot with logistic regression decision boundary (breast cancer data)?

How to explain low performance of naive Bayes on a dataset

disadvantages variational inference

The results of CV on Ridge are different than the results of RidgeCV

Does modeling with Random Forests require cross-validation?

Is one-hot encoding required for categorical variables in R (logistic regression)?