I'm having my head buried in the
AdaBoost.M1. Are there yet some variations of
AdaBoost.M1? I ask this because I read variations of algorithm listing from different sources.
For example, the original paper Experiments with a New Boosting Algorithm gives the following algorithm listing (screenshot attached):
In Chapter 10 of The Elements of Statistical Learning pp.339 the algorithm listing is gives as:
Also reading through Zhi-Hua Zhou (2012). Ensemble Methods Foundations and Algorithms, the author has this about the algorithm (pp. 54):
AdaBoost.M1 [Freund and Schapire, 1997] is a very straightforward extension, which is the same as the algorithm shown in Figure 2.2 except that the base learners now are multiclass learners instead of binary classifiers.
I have to admit, I'm confuse about this. Can someone explain this variations please?