Fdtxjr

How to run a prison with the smallest amount of guards?

How did Arya survive the stabbing?

How to safely derail a train during transit?

How do I extract a value from a time formatted value in excel?

Why are there no referendums in the US?

How can I kill an app using Terminal?

Detecting if an element is found inside a container

How does Loki do this?

Do the temporary hit points from Reckless Abandon stack if I make multiple attacks on my turn?

How to be diplomatic in refusing to write code that breaches the privacy of our users

What does "I’d sit this one out, Cap," imply or mean in the context?

What is the difference between "behavior" and "behaviour"?

Is there a problem with hiding "forgot password" until it's needed?

How to write papers efficiently when English isn't my first language?

How do scammers retract money, while you can’t?

Balance Issues for a Custom Sorcerer Variant

Is expanding the research of a group into machine learning as a PhD student risky?

What can we do to stop prior company from asking us questions?

Gears on left are inverse to gears on right?

Type int? vs type int

Is a stroke of luck acceptable after a series of unfavorable events?

What is the intuitive meaning of having a linear relationship between the logs of two variables?

Where does the Z80 processor start executing from?

Why escape if the_content isnt?

What is saturated model used in glm in R?

Binary Logistic Regression Model ProcessingWhen to use likelihood ratio test?Logistic regression with interactionsLogistic regression MLE example. What is this “logistic function”?Trouble analysing data set in RCalculate distribution of mean and variance given Gaussian data pointsModeling discrete data with a skew normal distribution and associated statistical testingProblem on the Interpretation of Specfic Regression Equation VariablesGLM model not corresponding to exploratory analysis (R-studio)Linear Regression Assumption: Normality of residual vs normality of variables

1

$begingroup$

So, I've read on stack exchange (glm summary explained) that residual deviance computed in the glm output is just the likelihood ratio chi-square stat comparing the saturated model to the reduced model. 1) But what saturated model is used? 2) How is the log likelihood evaluated at mle estimates computed for the given saturated model?

For simplicity, let's assume the glm is from the normal family and is just standard linear regression.

1)As I understand it a saturated model has as many parameters as data points-so there should be several (infinite) possible saturated models. For example, suppose you have two explanatory variables with four data points. One possible saturated model could be $y=b_1x_1+b_2x_2+b_3x_1x_2+b_4$ or another possible saturated model could be $y=b_1x_1+b_2x_2+b_3x_2^2+b_4$.

2)Regardless of the saturated model wouldn't the mle estimates for these coefficients just be found by solving the system directly using these four data points (as R naturally computes them)? Indeed, these would lead you to an infinite likelihood as residuals would be 0.

statistics linear-regression

share|cite|improve this question

edited Mar 22 at 20:24

Winston

asked Mar 17 at 20:47

WinstonWinston

10910

$endgroup$

add a comment | 

1

$begingroup$

So, I've read on stack exchange (glm summary explained) that residual deviance computed in the glm output is just the likelihood ratio chi-square stat comparing the saturated model to the reduced model. 1) But what saturated model is used? 2) How is the log likelihood evaluated at mle estimates computed for the given saturated model?

For simplicity, let's assume the glm is from the normal family and is just standard linear regression.

1)As I understand it a saturated model has as many parameters as data points-so there should be several (infinite) possible saturated models. For example, suppose you have two explanatory variables with four data points. One possible saturated model could be $y=b_1x_1+b_2x_2+b_3x_1x_2+b_4$ or another possible saturated model could be $y=b_1x_1+b_2x_2+b_3x_2^2+b_4$.

2)Regardless of the saturated model wouldn't the mle estimates for these coefficients just be found by solving the system directly using these four data points (as R naturally computes them)? Indeed, these would lead you to an infinite likelihood as residuals would be 0.

statistics linear-regression

share|cite|improve this question

edited Mar 22 at 20:24

Winston

asked Mar 17 at 20:47

WinstonWinston

10910

$endgroup$

add a comment | 

$begingroup$

So, I've read on stack exchange (glm summary explained) that residual deviance computed in the glm output is just the likelihood ratio chi-square stat comparing the saturated model to the reduced model. 1) But what saturated model is used? 2) How is the log likelihood evaluated at mle estimates computed for the given saturated model?

For simplicity, let's assume the glm is from the normal family and is just standard linear regression.

1)As I understand it a saturated model has as many parameters as data points-so there should be several (infinite) possible saturated models. For example, suppose you have two explanatory variables with four data points. One possible saturated model could be $y=b_1x_1+b_2x_2+b_3x_1x_2+b_4$ or another possible saturated model could be $y=b_1x_1+b_2x_2+b_3x_2^2+b_4$.

2)Regardless of the saturated model wouldn't the mle estimates for these coefficients just be found by solving the system directly using these four data points (as R naturally computes them)? Indeed, these would lead you to an infinite likelihood as residuals would be 0.

statistics linear-regression

share|cite|improve this question

edited Mar 22 at 20:24

Winston

asked Mar 17 at 20:47

WinstonWinston

10910

$endgroup$

share|cite|improve this question

edited Mar 22 at 20:24

Winston

asked Mar 17 at 20:47

WinstonWinston

10910

share|cite|improve this question

edited Mar 22 at 20:24

Winston

asked Mar 17 at 20:47

WinstonWinston

10910

asked Mar 17 at 20:47

WinstonWinston

10910

asked Mar 17 at 20:47

WinstonWinston

10910