This document discusses different machine learning classification algorithms that will be covered in the course, including K-Nearest Neighbors, Perceptron, Naive Bayes, Logistic Regression, Support Vector Machines, Decision Trees, and Neural Networks. It then focuses on Naive Bayes classifiers, explaining that they are simple, fast algorithms suitable for high-dimensional datasets like text classification, and that they find the best distribution to differentiate categories rather than separating data with a hyperplane like a perceptron.
This document discusses different machine learning classification algorithms that will be covered in the course, including K-Nearest Neighbors, Perceptron, Naive Bayes, Logistic Regression, Support Vector Machines, Decision Trees, and Neural Networks. It then focuses on Naive Bayes classifiers, explaining that they are simple, fast algorithms suitable for high-dimensional datasets like text classification, and that they find the best distribution to differentiate categories rather than separating data with a hyperplane like a perceptron.
This document discusses different machine learning classification algorithms that will be covered in the course, including K-Nearest Neighbors, Perceptron, Naive Bayes, Logistic Regression, Support Vector Machines, Decision Trees, and Neural Networks. It then focuses on Naive Bayes classifiers, explaining that they are simple, fast algorithms suitable for high-dimensional datasets like text classification, and that they find the best distribution to differentiate categories rather than separating data with a hyperplane like a perceptron.
This document discusses different machine learning classification algorithms that will be covered in the course, including K-Nearest Neighbors, Perceptron, Naive Bayes, Logistic Regression, Support Vector Machines, Decision Trees, and Neural Networks. It then focuses on Naive Bayes classifiers, explaining that they are simple, fast algorithms suitable for high-dimensional datasets like text classification, and that they find the best distribution to differentiate categories rather than separating data with a hyperplane like a perceptron.
Learning: ‘Bridging the Skills Gap’ Lesson 6: ‘Naïve’ Bayes Classier
We will look at a bunch of classiers in this course including:
Classication Coding Exercise Comment
Algorithm (Yes/No) K-Nearest Neighbors (KNN) ÷ Perceptron – Multi- x Perceptron is the Layer Perceptron simplest form of a Neural Network. Naïve Bayes ÷ Logistics Regression ÷ Support Vector Machines and Kernel ÷ Functions Decision Trees ÷ Neural Networks ÷ Naive Bayes models are fast and simple classication algorithms that are often suitable for very high-dimensional datasets. One application of naïve Bayes is text classication, aiming to assign documents (emails, tweets, posts, news) to one or many categories. One example is email spam/not- spam classication. The idea of naïve Bayes is that spam and not-span emails have a di5erent probability distributions.
For naïve Bayes the data doesn’t need to be linearly separable.
Naïve Bayes nds the hyperplane that best di5erentiates one distribution from another…it DOES NOT nd hyperplane that separates + from – A perceptron separates the data and will not converge when data is not linearly separable