User kjetil b halvorsen

120 votes

Accepted

What is the intuition behind SVD?

answered Oct 28, 2015 at 13:54

115 votes

Intuition on the Kullback–Leibler (KL) Divergence

answered Jan 7, 2016 at 21:42

89 votes

Accepted

How does saddlepoint approximation work?

answered Jan 21, 2016 at 14:03

71 votes

What is the difference between finite and infinite variance

answered May 27, 2014 at 12:04

59 votes

How does linear regression use the normal distribution?

answered Apr 29, 2015 at 8:07

58 votes

Accepted

Why should we use t errors instead of normal errors?

answered Oct 20, 2014 at 17:24

51 votes

Dropping one of the columns when using one-hot encoding

answered Feb 18, 2018 at 15:06

45 votes

Goodness of fit and which model to choose linear regression or Poisson

answered Mar 18, 2015 at 19:01

45 votes

Why do we only see $L_1$ and $L_2$ regularization but not other norms?

answered Mar 23, 2017 at 17:55

44 votes

Why does logistic regression become unstable when classes are well-separated?

answered Jan 2, 2017 at 18:16

43 votes

Accepted

What kind of information is Fisher information?

answered Feb 19, 2016 at 16:32

43 votes

Accepted

Correlations between continuous and categorical (nominal) variables

answered Jun 10, 2014 at 11:19

40 votes

Accepted

What is the intuition behind defining completeness in a statistic as being impossible to form an unbiased estimator of $0$ from it?

answered Feb 19, 2016 at 13:19

39 votes

Statistical interpretation of Maximum Entropy Distribution

answered Nov 10, 2016 at 11:02

35 votes

Accepted

Why Lasso or ElasticNet perform better than Ridge when the features are correlated

answered Feb 26, 2017 at 12:32

32 votes

The difference between with or without intercept model in logistic regression

answered Feb 6, 2017 at 9:50

30 votes

Family of GLM represents the distribution of the response variable or residuals?

answered Oct 30, 2018 at 15:55

29 votes

What is the difference between moment generating function and probability generating function?

answered Aug 13, 2017 at 14:20

29 votes

Accepted

What does "likelihood is only defined up to a multiplicative constant of proportionality" mean in practice?

answered May 13, 2014 at 15:12

29 votes

Taleb and the Black Swan

answered Sep 10, 2012 at 19:31

28 votes

Why do we need sigma-algebras to define probability spaces?

answered Mar 1, 2016 at 14:42

28 votes

Accepted

Can you give a simple intuitive explanation of IRLS method to find the MLE of a GLM?

answered Sep 28, 2016 at 15:17

27 votes

Is there a statistical application that requires strong consistency?

answered Nov 1, 2013 at 20:34

26 votes

Accepted

ML estimate of exponential distribution (with censored data)

answered Jan 14, 2015 at 11:27

26 votes

Accepted

Why is sum of squared residuals non-increasing when adding explanatory variable?

answered Aug 19, 2015 at 8:27

26 votes

Accepted

What are the consequences of "copying" a data set for OLS?

answered Jun 2, 2016 at 17:49

25 votes

Accepted

When to normalize data in regression?

answered Mar 16, 2016 at 14:03

24 votes

Expected value as a function of quantiles?

answered Aug 5, 2015 at 12:31

24 votes

Accepted

Why doesn't the exponential family include all distributions?

answered Jul 31, 2017 at 10:41

23 votes

What is meant by a "random variable"?

answered Feb 26, 2017 at 10:26

Stack Exchange Network

2,553 Answers

What is the intuition behind SVD?

Intuition on the Kullback–Leibler (KL) Divergence

How does saddlepoint approximation work?

What is the difference between finite and infinite variance

How does linear regression use the normal distribution?

Why should we use t errors instead of normal errors?

Dropping one of the columns when using one-hot encoding

Goodness of fit and which model to choose linear regression or Poisson

Why do we only see $L_1$ and $L_2$ regularization but not other norms?

Why does logistic regression become unstable when classes are well-separated?

What kind of information is Fisher information?

Correlations between continuous and categorical (nominal) variables

What is the intuition behind defining completeness in a statistic as being impossible to form an unbiased estimator of $0$ from it?

Statistical interpretation of Maximum Entropy Distribution

Why Lasso or ElasticNet perform better than Ridge when the features are correlated

The difference between with or without intercept model in logistic regression

Family of GLM represents the distribution of the response variable or residuals?

What is the difference between moment generating function and probability generating function?

What does "likelihood is only defined up to a multiplicative constant of proportionality" mean in practice?

Taleb and the Black Swan

Why do we need sigma-algebras to define probability spaces?

Can you give a simple intuitive explanation of IRLS method to find the MLE of a GLM?

Is there a statistical application that requires strong consistency?

ML estimate of exponential distribution (with censored data)

Why is sum of squared residuals non-increasing when adding explanatory variable?

What are the consequences of "copying" a data set for OLS?

When to normalize data in regression?

Expected value as a function of quantiles?

Why doesn't the exponential family include all distributions?

What is meant by a "random variable"?