This document summarizes different types of learning methods in neural networks. It discusses multilayer feedforward neural networks, which have an input layer, hidden layer, and output layer connected by weights. The document then categorizes common neural network learning algorithms into supervised learning (including gradient descent and stochastic learning), unsupervised learning (such as Hebbian and competitive learning), and reinforced learning. Specific learning methods like backpropagation, Hebbian rule, and Boltzmann learning are also overviewed.