User Max Gordon

153 votes

5 answers

212k views

Why normalize images by subtracting dataset's image mean, instead of the current image mean in deep learning?

asked May 8, 2016 at 11:11

51 votes

3 answers

17k views

Are splines overfitting the data?

asked Feb 1, 2013 at 9:36

20 votes

3 answers

19k views

Time dependent coefficients in R - how to do it?

asked Nov 18, 2011 at 9:13

17 votes

3 answers

14k views

How to best display graphically type II (beta) error, power and sample size?

asked Aug 11, 2011 at 7:49

15 votes

4 answers

3k views

How to find out if an online poker-site is fair?

asked May 27, 2012 at 10:15

12 votes

1 answer

7k views

Logistic quantile regression – how to best convey the results

asked May 5, 2012 at 10:35

11 votes

1 answer

15k views

Different prediction plot from survival coxph and rms cph

asked Jan 30, 2012 at 18:37

9 votes

2 answers

28k views

How to do ROC-analysis in R with a Cox model

asked Oct 24, 2011 at 14:03

9 votes

1 answer

552 views

Restricting minimum subgroup size in a bootstrap resampling study - why is this approach wrong?

asked Sep 22, 2013 at 12:36

6 votes

1 answer

707 views

Does inference from a heteroskedasticity-consistent covariance matrix follow the t-distribution or the normal?

asked Sep 8, 2013 at 13:21

6 votes

2 answers

584 views

Risk assesment and non-statistician's perception of percentages

asked Jun 9, 2012 at 9:12

6 votes

3 answers

2k views

Is using a questionnaire score (EuroQol's EQ-5D) with a bimodal distribution as outcome in linear regression a problem?

asked Feb 9, 2012 at 1:56

6 votes

1 answer

1k views

Assumptions and pitfalls in competing risks model

asked Aug 6, 2011 at 10:09

5 votes

1 answer

1k views

Combining individual terms from the predict() function

asked Dec 27, 2015 at 13:25

3 votes

1 answer

4k views

Poisson regression in a survival setting on a simulated set

asked Sep 3, 2011 at 17:49

2 votes

3 answers

3k views

Competing risk regression with CRR slow on large datasets?

asked Jul 17, 2011 at 9:31

2 votes

1 answer

743 views

How to get forestplot to accept an expression in row label?

asked Aug 16, 2011 at 17:05

2 votes

1 answer

75 views

Testing subgroups for significance in cox regression - how to format data

asked Dec 21, 2015 at 15:00

1 vote

0 answers

2k views

Finding the maximum contributer variables by applying the SVD - should the absolute values be used or just the maximum value?

asked Feb 27, 2013 at 11:20

1 vote

0 answers

58 views

Pooling estimates and variances with zero counts

asked Nov 27, 2018 at 11:48

0 votes

0 answers

44 views

What do you call a weighted summary statisitc?

asked Feb 20, 2020 at 18:09

Stack Exchange Network

21 Questions

Why normalize images by subtracting dataset's image mean, instead of the current image mean in deep learning?

Are splines overfitting the data?

Time dependent coefficients in R - how to do it?

How to best display graphically type II (beta) error, power and sample size?

How to find out if an online poker-site is fair?

Logistic quantile regression – how to best convey the results

Different prediction plot from survival coxph and rms cph

How to do ROC-analysis in R with a Cox model

Restricting minimum subgroup size in a bootstrap resampling study - why is this approach wrong?

Does inference from a heteroskedasticity-consistent covariance matrix follow the t-distribution or the normal?

Risk assesment and non-statistician's perception of percentages

Is using a questionnaire score (EuroQol's EQ-5D) with a bimodal distribution as outcome in linear regression a problem?

Assumptions and pitfalls in competing risks model

Combining individual terms from the predict() function

Poisson regression in a survival setting on a simulated set

Competing risk regression with CRR slow on large datasets?

How to get forestplot to accept an expression in row label?

Testing subgroups for significance in cox regression - how to format data

Finding the maximum contributer variables by applying the SVD - should the absolute values be used or just the maximum value?

Pooling estimates and variances with zero counts

What do you call a weighted summary statisitc?