This class describes the object that will take the objects of the Dataset type where data were fit with different models, and compare the results.
NOTE: One theoretical aspect that needs to be clarified in the original publications is that here we are using WEIGHTED sum of squares instead of just sum of squares as asked by the the original definition of the test. Intuitively, it should not make difference (say: reverse results, etc) because both compared models utilize the same formula for sum of squares. However, the sensitivity of the test may be altered so be cautious with interpretation until the issue is clarified.
To compute AICc run compute_AICc(dataset) method. The method returns a text report and AICc value.
To compare results of fitting of the same dataset with two models first create a Dataset object with the data and then copy it using its make_a_copy() method. Then set different models there and fit the data. Now call compare_datasets() method with these two datasets as parameters.
If you need to compare more than two datasets at a time use methods add_dataset() and evaluate_datasets(). Use add_dataset() method to load a dataset handle into the array. It is your duty to make sure you a comparing 'apples to apples', that is you load a copy of the same dataset fitted with different models, not the datasets of different type!
When you loaded dataset you simply created the link. You don't need to reload it with new fitting results. Method evaluate_datasets() goes through the array and computes AICc for all of them the model with lowest AICc is the most probable.
For more information see DatasetAIC.m