User Funkwecker

77 votes

7 answers

190k views

Which activation function for output layer?

asked Jun 12, 2016 at 14:42

18 votes

2 answers

24k views

Number of parameters in an artificial neural network for AIC

asked Sep 28, 2015 at 16:24

15 votes

3 answers

2k views

Deep neural networks -- Just for image classification?

asked Nov 18, 2015 at 12:07

14 votes

1 answer

6k views

AIC of ridge regression: degrees of freedom vs. number of parameters

asked Dec 2, 2016 at 11:34

13 votes

1 answer

12k views

Is bias the same as underfitting and variance the same as overfitting?

asked Dec 9, 2016 at 8:27

11 votes

3 answers

17k views

Normalization of dummy variables

asked Apr 25, 2014 at 7:05

10 votes

4 answers

10k views

Is data-driven modelling and machine learning the same thing?

asked Jan 27, 2021 at 8:03

9 votes

2 answers

15k views

Should k-means only be applied if the variables are normally distributed?

asked Jan 29, 2018 at 17:32

8 votes

1 answer

2k views

How can I calculate the AICc if the number of samples equals the number of parameters plus one

asked Nov 29, 2016 at 11:10

8 votes

2 answers

1k views

What is the disadvantage of repeated cross-validation?

asked Feb 4, 2020 at 16:30

8 votes

1 answer

7k views

Which samples are used in random forests for calculating variable importance?

asked May 18, 2015 at 9:26

7 votes

3 answers

4k views

Does it make sense to use dynamic time warping when clustering time series that all have the same length and sampling interval?

asked Jul 17, 2013 at 12:33

7 votes

1 answer

3k views

Comparing differences of AIC of different data sets

asked Dec 12, 2016 at 8:40

7 votes

2 answers

716 views

Which statistic to report for repeated cross-validation?

asked Feb 8, 2020 at 7:04

6 votes

0 answers

3k views

When should I use AIC instead of cross-validation?

asked Mar 14, 2016 at 10:31

5 votes

1 answer

2k views

Mantel test vs. Pearson's correlation coefficient

asked Nov 20, 2012 at 13:41

5 votes

1 answer

3k views

Does using lagged independent variables makes sense?

asked Jan 5, 2016 at 8:37

5 votes

2 answers

14k views

Is AIC a measure of goodness of fit? [duplicate]

asked Apr 1, 2017 at 16:30

5 votes

0 answers

244 views

How to test if coefficients of two linear regressions differ in practice?

asked Aug 9, 2016 at 6:02

5 votes

1 answer

2k views

Why not using cross validation for estimating the error of a linear model?

asked Nov 3, 2014 at 11:35

5 votes

3 answers

651 views

Why does the rank order of models differ for R squared and RMSE?

asked Apr 22, 2018 at 16:46

5 votes

1 answer

2k views

Does it make sense to compare models using a paired t-test?

asked Mar 9, 2018 at 15:31

4 votes

2 answers

76 views

How to quantify the reliablility of an estimate

asked Oct 24, 2017 at 9:05

4 votes

1 answer

668 views

Difference of variable selection and importance estimation

asked Dec 15, 2014 at 10:06

4 votes

0 answers

215 views

Are there any good reasons for applying Bayesian Networks these days?

asked Jul 20, 2015 at 11:58

4 votes

1 answer

916 views

What is the number of parameters for AIC if some coefficients are zero?

asked Nov 14, 2016 at 8:52

4 votes

0 answers

2k views

Using AIC or cross-validated MSE for selecting neural network models for time series prediction

asked Jan 4, 2016 at 14:35

3 votes

1 answer

6k views

Testing significance of RMSE of models

asked Dec 4, 2015 at 10:15

3 votes

0 answers

418 views

What is the number of parameters of an elastic net model?

asked Nov 28, 2016 at 6:08

3 votes

2 answers

1k views

Model selection using mean AIC for very huge data sets

asked Sep 5, 2014 at 10:24

Stack Exchange Network

66 Questions

Which activation function for output layer?

Number of parameters in an artificial neural network for AIC

Deep neural networks -- Just for image classification?

AIC of ridge regression: degrees of freedom vs. number of parameters

Is bias the same as underfitting and variance the same as overfitting?

Normalization of dummy variables

Is data-driven modelling and machine learning the same thing?

Should k-means only be applied if the variables are normally distributed?

How can I calculate the AICc if the number of samples equals the number of parameters plus one

What is the disadvantage of repeated cross-validation?

Which samples are used in random forests for calculating variable importance?

Does it make sense to use dynamic time warping when clustering time series that all have the same length and sampling interval?

Comparing differences of AIC of different data sets

Which statistic to report for repeated cross-validation?

When should I use AIC instead of cross-validation?

Mantel test vs. Pearson's correlation coefficient

Does using lagged independent variables makes sense?

Is AIC a measure of goodness of fit? [duplicate]

How to test if coefficients of two linear regressions differ in practice?

Why not using cross validation for estimating the error of a linear model?

Why does the rank order of models differ for R squared and RMSE?

Does it make sense to compare models using a paired t-test?

How to quantify the reliablility of an estimate

Difference of variable selection and importance estimation

Are there any good reasons for applying Bayesian Networks these days?

What is the number of parameters for AIC if some coefficients are zero?

Using AIC or cross-validated MSE for selecting neural network models for time series prediction

Testing significance of RMSE of models

What is the number of parameters of an elastic net model?

Model selection using mean AIC for very huge data sets