Deviance Residuals

🐈🐈cats too can watch statquesttt 😀. Nice explanation

Mr Josh, if I used Iteratively reweighted least squares to get the parameters. Can I use the theDeviance in measure of goodness of fit a model?! I would be grateful if you could explain Iteratively reweighted 🙏🏻🙏🏻
Thanks in advance.

Mr Josh, I have been considering the intuition of deviance. Isn't it a bad thing to compare a proposed model to an overfitted model even if we can determine that the proposed model statistic isn't significantly different from the overfitted model i.e making it a "good model". Isn't it better to use cross-validation methods. I only ask because you also teach machine learning methods. I have been trying to do cross-validation techniques to GLMS don't know if deviance may be better.

I love this song!

Awesome video as always!

Hi Josh! Is there a paper of yours (or one that could please recommend), which I can cite regarding the use of residual deviance for logistic goodness of fit? Thanks! Andy

I feel like I'm watching Blue Clues.. Helpful nonetheless

Thank you sir!
This is really help me a lot!

I had the same confusion when I first notice a value>1 when calculating likelihood after just watching the video “logistic regression details Pt2 Maximum likelihood”. Then I realized the video “saturated model and deviance” and this video are explained using normal distribution, and the likelihood here are prob density calculated using pdf of normal dist, which can be >1

Excellent video. Could you please create a specific video on "Quasi Poisson Regression"

...the intro.... that.. intro.... IT'S ABSOLUTELY TRUE for me! Made my day

That is a bam. Thank you.

Can we use this as a substitute for regular residuals? I am working on a probit model. The command predict resid, residuals gives me an error but works perfectly in case of an OLS reg. How do I check the residuals in case of a probit model?

Why does the likelihood get maximized when we have a saturated model? Does a saturated model have multiple MLEs?

In GLM, what is the difference between, deviance and residual deviance (in simple term), pls reply

i want to be a cat 😩

Hi Josh,
How did deviance turn into square root?

How these likelihoods are greater than 1...... can't understand that part...... you didn't mention how you calculated it

Why in the case of logisitic regression you do not calculate the likelihood of the points to get the deviance? Can you relate the deviance test to the Hosmer-Lemeshow test?

Hi Josh,
While I understand the definitions, I don't get why the individual contributing deviances are called "Deviance Residual". It's quite confusing especially there is already a Residual Deviance. I guess this is more like an English than a math question - could you please enlighten?

I'm not a cat but I do sleep on a couch :(

the way you explain every concept so clearly is just awesome ❤❤❤❤

To plainly state formulas is truly an art and gift. Thankyou

Why the degree of freedom of chi-square is the difference between null and proposal?

Use these to identify outliers in the data