The Expectation-Maximization (EM) in theoretical details
NOTE: This blog post consists of images. It might take a while to load in your browser!
Support Vector Machines in theoretical details
NOTE: This blog post consists of images. It might take a while to load in your browser! Visual Representation of Kernels A kernelized SVM is equivalent to a linear SVM...
Can you interpret probabilistically the output of a Support Vector Machine?
SVMs don’t output probabilities natively, but probability calibration methods can be used to convert the output to class probabilities. Various methods exist, including Platt scaling (particularly suitable for SVMs) and...
Univariate/Multivariate Gaussian Distribution and their properties
Univariate Normal Distribution The normal distribution, also known as Gaussian distribution, is defined by two parameters, mean $\mu$, which is expected value of the distribution and standard deviation $\sigma$ which...
Softmax Regression
Softmax Regression (a.k.a. Multinomial Logistic, Maximum Entropy Classifier, or just Multi-class Logistic Regression) is a generalization of logistic regression that we can use for multi-class classification (under the assumption that...