Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

The following figure shows the log-likelihood of individual observations for different subsets of a data set - the horizontal line shows the average log-likelihood:

figure one

So, the first subset includes all the data points observed up to t=7, the second all data points up to t=33, and so forth. For each subset, I estimate a parameter $\theta$, and then I calculate the likelihood of each observation given this parameter estimate. This is what the figure shows.

It can be seen that on average the likelihood of the observed data (shown as a horizontal line) tends to decrease as the number of observations grows. This pattern can be confirmed in the following plot, which shows how the average likelihood of observations changes as the number of observations included grows, for different data sets:

enter image description here

In the light of this, would you agree that these results suggest that the $\theta$ parameter is not equal for all $t$? If it was equal, I would expect the same average likelihood regardless of the number of observations included.

share|improve this question
Could you explain why you are carrying out this analysis? Are you looking for outliers? Changepoints? (If so, what kind of change?) Something else? – whuber Nov 24 '12 at 23:29
Nevermind, the analysis is wrong. – Ernest A Nov 26 '12 at 1:04

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.