济南大学计算智能实验室陈月辉 15499.ppt

分布估计算法 思想 遗传算法中的交叉、变异等操作有可能破坏已经优化好的个体。为了避免这种现象，一种新的演化算法 – 分布估计算法应运而生。 分布估计算法中没有交叉和变异。主要用到是好的个体的一种概率模型，以及根据此模型抽样产生下一代。 GA to EDA 基于种群的增强式学习 Population based Incremental Learning (PBIL, Baluja, 1994) Basic PBIL P ? initialize probability vector (each position = 0.5) while (generations++ limit) for each vector i do for each position j do generate Vi(j) according to P(j) end-do evaluate f(Vi) end-do Vmax = max(f(Vi)) update P according to Vmax if random(0,1] Pmutate mutate P end-if end-while Details Population replaced by probability vector P = {p1 , p2 , …, p?} pi : Probability of 1 in the ith bit Generate n individuals Update P using the best individual pi(t+1) = ?xi + pi(t)(1- ?), i = 1,2,…,? Mutate P: pi(t+1) = ?mU[0,1) + pi(t+1)(1- ?m) PBIL Example t = 0, P = {0.5, 0.5, 0.5, 0.5} Generate 5 individuals {1010, 1100, 0100, 0111, 0001} Fitness: {2, 2, 1, 3, 1} Best individual: 0111; ? = 0.1 Update P p1 = 0.5*(1-0.1) = 0.45 p2 = p3 = p4 = 0.1*1 + 0.5*(1-0.1) = 0.55 Some applications Function optimization Job-shop scheduling TSP Bin-packing Knapsack Problem Neural Network weight training 分布估计算法：总的框架 Estimation of Distribution Algorithms do just that! Typically they operate as follows: Step 0: Randomly generate a set of ? individuals (t=0) Step 1: Evaluate the ? individuals While (not done) Step 2: Select ? individuals (where ? ? ?) to be parents Develop a probability distribution/density function, pt, based on the parents Step 3: Create ? offspring using pt Step 4: Evaluate the offspring Step 5: The ? offspring replace the ? parents (t = t + 1) Step 6: Goto While Flowchart What Models to use? Start with Probability vector for binary strings Gaussian distribution Later Dependency tree models (COMIT) Bayesian Network Probability Vector PMBGAs 分布估计算法：概率向量 The EDA is known as the Univariate Marginal Distribution Algorithm Let’s try to solve the following problem f(x) = x2, where -2.0 ? x ? 2.0, Let l = 7, therefore our mapping function will be d(2,-2,7,c) =