A Perceptron is an algorithm used for supervised learning of binary classifiers.
A Averaged Perceptron is an algorithm used for supervised learning of binary classifiers.
The average perceptron algorithm uses the same rule to update parameters but The final returning values of θ and θ₀ are the average of all the values of θ and θ₀ in each step.
A Pegasos is an algorithm used for supervised learning of binary classifiers.
The pegasos algorithm has the hyperparameter λ, giving more flexibility to the model to be adjusted.
Comparison between the convergence of perceptron, average perceptron, and pegasos linear classifiers on a toy dataset
Sentiment analysis of a food review dataset with bag-of-words model using pegasos classifier. Most positive and negative explainatory words are shown.
SVM sklearn cancer