User confused

23 votes

5 answers

4k views

Can someone explain the importance of mean stationarity in time series?

asked Jun 1, 2020 at 5:18

21 votes

2 answers

4k views

Can someone give the intuition behind Mean Absolute Error and the Median? [duplicate]

asked Sep 22, 2019 at 0:17

18 votes

4 answers

2k views

Does Bayesian statistics bypass the need for the sampling distribution?

asked May 22, 2020 at 8:17

12 votes

2 answers

812 views

What's the rationale for not checking residuals when building a ML model?

asked Jun 7, 2020 at 22:02

10 votes

3 answers

396 views

Would you say this is a trade off between frequentist and Bayesian stats?

asked May 22, 2020 at 8:38

10 votes

2 answers

4k views

Can someone explain to me the parameters of a lognormal distribution?

asked Sep 18, 2019 at 14:07

10 votes

1 answer

335 views

Does Lasso make irrelevant the need for coefficient significance testing?

asked Nov 6, 2020 at 20:01

7 votes

1 answer

1k views

Why does the likelihood function of a binomial distribution not include the combinatorics term? [duplicate]

asked Oct 14, 2019 at 1:25

6 votes

2 answers

3k views

In ML, once we remove a feature, can we safely assume that feature will not be important again?

asked Jun 8, 2020 at 10:01

6 votes

2 answers

3k views

Why do we not interpret main effects if interaction terms are significant in ANOVA?

asked Sep 7, 2020 at 9:23

5 votes

1 answer

310 views

Why in Hamiltonian MCMC do we multiply the posterior distribution by the likelihood?

asked Nov 7, 2019 at 17:52

4 votes

2 answers

2k views

Are all slope coefficients correlated with the intercept in multiple linear regression?

asked Nov 27, 2019 at 18:32

4 votes

1 answer

487 views

Do we use different number of dummy variables for Classical and Bayesian stats?

asked Dec 10, 2019 at 4:33

4 votes

1 answer

5k views

How do you interpret prediction output in GBM() in R for classification problem?

asked Feb 7, 2020 at 15:44

4 votes

3 answers

176 views

How to interpret $x \in \mathbb{R}^n$?

asked Mar 17, 2020 at 4:44

4 votes

1 answer

63 views

Typo in Deepleariningbook.org or am I misunderstanding Bayesian stats?

asked Mar 19, 2020 at 22:43

4 votes

1 answer

230 views

Reference for doing linear regression with mean absolute deviation?

asked Mar 20, 2020 at 2:15

4 votes

2 answers

139 views

Are time series models limited in real life application and are primarily used to model the residuals of another model?

asked Jun 5, 2020 at 21:07

3 votes

4 answers

4k views

What's the relationship between cointegration and linear regression?

asked Jun 27, 2020 at 18:34

3 votes

1 answer

178 views

Information loss in recurrent neural networks?

asked Jul 29, 2020 at 6:21

3 votes

1 answer

2k views

Why would Lasso recommend a lambda of 0?

asked Mar 4, 2021 at 5:01

3 votes

3 answers

3k views

What is a "shallow" layer and a "deep" layer of a neural network?

asked Jul 26, 2020 at 17:01

3 votes

1 answer

359 views

Should you exclude one level of a categorical variable when it's part of an interaction term?

asked Jul 28, 2021 at 21:35

3 votes

1 answer

229 views

How do you interpret LASSO dummy coefficients?

asked Sep 25, 2021 at 0:17

3 votes

1 answer

419 views

Does Maximum Likelihood Estimation solve n < p problem in regression?

asked May 22, 2020 at 9:42

3 votes

2 answers

725 views

Rationale for taking second derivative in least squares estimation?

asked May 23, 2020 at 23:12

3 votes

1 answer

652 views

Where is the divide between information criterion (AIC, BIC, etc...) and cross validation?

asked Feb 8, 2020 at 19:06

3 votes

1 answer

117 views

When would it be computationally inefficient to sample from a distribution?

asked Oct 30, 2019 at 4:05

3 votes

2 answers

184 views

What exactly is violated if your response variable exhibits autocorrelation in ordinary least squares regression?

asked Feb 16, 2020 at 17:57

3 votes

1 answer

245 views

Clarification on how decision trees split up categorical variables?

asked Jan 8, 2020 at 2:56

Stack Exchange Network

112 Questions

Can someone explain the importance of mean stationarity in time series?

Can someone give the intuition behind Mean Absolute Error and the Median? [duplicate]

Does Bayesian statistics bypass the need for the sampling distribution?

What's the rationale for not checking residuals when building a ML model?

Would you say this is a trade off between frequentist and Bayesian stats?

Can someone explain to me the parameters of a lognormal distribution?

Does Lasso make irrelevant the need for coefficient significance testing?

Why does the likelihood function of a binomial distribution not include the combinatorics term? [duplicate]

In ML, once we remove a feature, can we safely assume that feature will not be important again?

Why do we not interpret main effects if interaction terms are significant in ANOVA?

Why in Hamiltonian MCMC do we multiply the posterior distribution by the likelihood?

Are all slope coefficients correlated with the intercept in multiple linear regression?

Do we use different number of dummy variables for Classical and Bayesian stats?

How do you interpret prediction output in GBM() in R for classification problem?

How to interpret $x \in \mathbb{R}^n$?

Typo in Deepleariningbook.org or am I misunderstanding Bayesian stats?

Reference for doing linear regression with mean absolute deviation?

Are time series models limited in real life application and are primarily used to model the residuals of another model?

What's the relationship between cointegration and linear regression?

Information loss in recurrent neural networks?

Why would Lasso recommend a lambda of 0?

What is a "shallow" layer and a "deep" layer of a neural network?

Should you exclude one level of a categorical variable when it's part of an interaction term?

How do you interpret LASSO dummy coefficients?

Does Maximum Likelihood Estimation solve n < p problem in regression?

Rationale for taking second derivative in least squares estimation?

Where is the divide between information criterion (AIC, BIC, etc...) and cross validation?

When would it be computationally inefficient to sample from a distribution?

What exactly is violated if your response variable exhibits autocorrelation in ordinary least squares regression?

Clarification on how decision trees split up categorical variables?