The Adaboost method for creating a strong binary classifier from a series of weak classifiers is implemented in this project. We use decision stumps as our weak classifiers. Classification results are shown for some synthetic datasets and the MNIST dataset containing images of digits.
Refer question 1 from the file hw4.pdf
. All subparts have been implemented. Run /code/myMainScript.m
to generate the results for the subparts sequentially. The function descriptions are provided in the respective files.