User dsimcha

109 votes

12 answers

20k views

What, precisely, is a confidence interval?

asked Jan 28, 2011 at 0:23

57 votes

7 answers

34k views

Best PCA algorithm for huge number of features (>10K)?

asked Sep 18, 2010 at 2:08

47 votes

8 answers

11k views

Rigorous definition of an outlier?

asked Feb 13, 2011 at 15:07

15 votes

2 answers

1k views

Sane stepwise regression?

asked Jan 29, 2011 at 4:44

14 votes

3 answers

8k views

Best way to put two histograms on same scale?

asked Mar 3, 2011 at 16:28

11 votes

2 answers

792 views

Finding number of gaussians in a finite mixture with Wilks' theorem?

asked Oct 7, 2011 at 16:18

9 votes

3 answers

1k views

How to visualize the true dimensionality of the data?

asked Mar 9, 2011 at 17:49

7 votes

4 answers

4k views

Determining trend significance in a time series

asked Dec 5, 2010 at 2:33

7 votes

3 answers

1k views

How do extreme values scale with sample size?

asked Jan 25, 2011 at 4:16

7 votes

2 answers

1k views

Big-O Scaling of R's cmdscale()

asked May 17, 2011 at 14:16

5 votes

1 answer

504 views

Estimating Poisson waiting times: using the last, censored sample?

asked Mar 30, 2012 at 1:26

3 votes

1 answer

667 views

Confidence Intervals for Holdout R^2?

asked Mar 18, 2011 at 21:04

2 votes

2 answers

541 views

How does required sample size scale with the number of features in multiple linear regression?

asked Dec 23, 2011 at 21:40

Stack Exchange Network

13 Questions

What, precisely, is a confidence interval?

Best PCA algorithm for huge number of features (>10K)?

Rigorous definition of an outlier?

Sane stepwise regression?

Best way to put two histograms on same scale?

Finding number of gaussians in a finite mixture with Wilks' theorem?

How to visualize the true dimensionality of the data?

Determining trend significance in a time series

How do extreme values scale with sample size?

Big-O Scaling of R's cmdscale()

Estimating Poisson waiting times: using the last, censored sample?

Confidence Intervals for Holdout R^2?

How does required sample size scale with the number of features in multiple linear regression?