Finally, the last classifier. Will it be the best one at the end? Yes! Previous SVM classification will be replaced by a discriminant analysis model. Discriminant analysis in short: its principle is very similar to the naive Bayes approach – probabilities, statistics and optimization mixed up together. The approach is Read more…
Our journey to find the best classifier is not yet at the end. The last introduced classifier was the naive Bayes and now the next one – support vector machine itself – has a difficult task to win this competition. The naive classifier set a high bar. Spoiler: the support Read more…
Same task, but third classification algorithm – naive Bayes. I was looking forward, because a MATLAB script was already written and I naively thought my only work is to replace one function name (fitcknn to fitcnb). My good mood lasted until I realized that optimization of hyperparameters did not have Read more…
The aim of this project is still the same, only a classification algorithm will be different. No more trees anymore. They will be replaced by nearest neighbors, but the method is the same as previously. K-nearest neighbors (k-NN) is a simple classification algorithm, which assign a class to an item Read more…
Warning: this story has no happy ending. In the previous post I promised that we will grow better trees. For these kinds of problems optimization is used. In MATLAB, function fitctree which returns fitted classification tree can have an additional argument ‘OptimizeHyperparameters’ with value ‘auto’ or ‘all’ (and also ‘none’, Read more…