[AIM] Evolutionary Algorithms II

Day 5

Outline

  • Generic permutation based operators
    • Partially mapped crossover(PMX)
    • Order crossover (OX)
    • Cycle crossover (CX)
  • Memetic Algorithms
  • Multimeme (self-adaptive/self-generating) memetic algorithms

Issues with binary representation in GA

Binary representation and classical crossover and mutation operators are not suitable for encoding permutation
for TSP, could cause illegal tours that:
- not all cities visited
- cities visited more than once
- loops formed within the tour
Because it is a permutation, there should not have any number that overlap.

So how do we solved this problem?

Partially mapped crossover

  • Choose two random cut points to serve as swapping boundaries
  • Swap segments between cut points
  • Preserving the order and position of as many cities as possible from other parent

Order Crossover (OX)

  • Choosing a subsequence of a tour from one parent
  • Preserving relative order of cities from other parent

Cycle Crossover (CX)


Memetic Algorithms(Population based metaheuristic)

  • Meme: Contagious piece of information, it is like culture, information sharing
  • Memes are similar to local refinement/local search
  • Gene vs Meme: Memes can change, evolve using rules and time scales other than the traditional genetic ones.
  • Aims to improve GAs by embedding local search
  • MAs make uses of exploration capabilities of GAs and exploitation capabilities of local search
  • MAs have an explicit mechanism to balance exploitation and exploration
  • Memetic Algorithms shown to be much faster and more accurate than GAs on some problems.
  • It is a traditional way of applying MAs. For local search, you also can apply it after initialisation, after crossover, after mutation.
When we apply for one algorithm, meme is not only local search, but maybe no meme(GA), maybe some other local search.

Note: Different memes yield different performances, designing the right meme is important

Multimeme memetic algorithms

Now meme is not only just local search!
  • Introduce memetic material for each individuals
  • A meme encodes how to apply an operator(which one to apply, max number of iterations, acceptance criteria), when to apply, where to apply, how frequent to apply.
  • Meme of each operator can be combined under a memeplex

MMA features

  • Memes represent instructions for self-improvement, specify set of rules, programs, heuristics, strategies, behaviors, etc.
  • Interaction between memes and genes are not direct, mediated by individual
  • Memes can evolve, change by means of non-traditional genetic transformations and metrics

Mutating memes during evolution

You can mutate memes during evolution by innovation rate.
  • Innovation rate (IR) is in the range of [0,1], which is the probability of mutating a meme
  • If the random number is less than IR, it will mutate the meme.
  • For example, for a MAX-SAT problem. IR is 0.5 and now it is ( 1 0 0 0 1+ HC2). Then we random a number, the number is 0.3 < 0.5, so we mutate it. It might be ( 1 0 0 0 1 + HC0)
  • if IR = 0, no innovation. If an allele is not introduced in the initial generation, it will not be reintroduced again.
  • if IR = 1, everything will be mutated. all different strategies implied by the available M memes might be equally used

Simple Inheriting Mechanism(SIM)


For example:
For a MAX-SAT problem:
(1 0 | 0 0 1 + HC1 HS1 L2) f =2
(1 1 | 0 1 1 + HC0 HS4 L1) f =1
“|” means the crossover point
Because the second one fitness value is lower, so we apply its memeplex. (1 1 0 0 1 + HC0 HS4 L1)
(1 0 0 1 1 + HC0 HS4 L1)

评论

发表评论

此博客中的热门博文

[MLE] Linear Classification

图片
Linear Classification Discriminant functions Least squares for classification Fisher’s linear discriminant K-Nearest Neighbour(KNN) Classification What’s decision boundaries/linearly separable? The goal in classification is to take an input vector x and to assign it to one of K discrete classes \(C_k\) where k = 1,…,K. In the most common scenario, the classes are taken to be disjoint, so that each input is assigned to one and only one class. The input space is thereby divided into decision regions whose boundaries are called decision boundaries or decision surfaces. In this blog, we consider linear models for classification , by which we mean that the decision surfaces are linear functions of the input vector x and hence are defined by (D-1) -dimensional hyperplanes within the D-dimensional input space. For example: in 3D input space, the decision boundary will be a 2D surface What is Linear Separability Linearly separable data: Datasets whose classes can be separated by
阅读全文

[AIM] MetaHeuristics

图片
Day2 Outline Metaheuristics -definition Single point based search method(simple stochastic local search) Iterated local search Scheduling Tabu Search Metaheuristic what is a metaheuristic? A metaheuristic is a high-level problem independent algorithmic framework that provides a set of guidelines or strategies to develop heuristic optimization algorithms. what is the difference between heuristic and metaheuristic? Heuristic : Heuristics are problem-dependent techniques. As such, they usually are adapted to the problem at hand and they try to take full advantage of the particularities of this problem. However, because they are often too greedy, they usually get trapped in a local optimum and thus fail, in general, to obtain the global optimum solution. Metaheuristic : Meta-heuristics, on the other hand, are problem-independent techniques . As such, they do not take advantage of any specificity of the problem and, therefore, can be used as black boxes . In general, t
阅读全文

[CS231] Neural Networks

图片
Neural Network Table of Contents: Quick intro without brain analogies Modeling one neuron Biological motivation and connections Single neuron as a linear classifier Commonly used activation functions Neural Network architectures Layer-wise organization Summary Additional references Quick intro It is possible to introduce neural networks without appealing to brain analogies. In the section on linear classification we computed scores for different visual categories given the image using the formula \( s = W x \), where \(W\) was a matrix and \(x\) was an input column vector containing all pixel data of the image. In the case of CIFAR-10, \(x\) is a [3072x1] column vector, and \(W\) is a [10x3072] matrix, so that the output scores is a vector of 10 class scores. An example neural network would instead compute \( s = W_2 \max(0, W_1 x) \). Here, \(W_1\) could be, for example, a [100x3072] matrix transforming the image into a 100-dimensional intermediate vector. The funct
阅读全文