 Today, we are going to discuss the topic genetic algorithms. At the end of this session, students will be able to demonstrate the working of a genetic algorithm and be acquainted with its operators to be used for randomized search. Genetic algorithm is a part of evolutionary computation, where every individual is called as a chromosome, it has hereditary characteristics and they are coded replications of characters called as genes which are present on this chromosomes. Every chromosome is an array of characters and each character determines a gene. GA is population-based, where a population is a combination of many chromosomes and it is an optimized technique. GA follows biological principles of Darwin's theory based on natural search and genetics. It combines the survival of the fittest among string structures with a structural yet randomized information exchange. It is a robust search in a complex space and works with coding of parameter sets. We use payoff objective functions to determine whether a chromosome is fitter or is not going to be used for the next generation. Everywhere we use a probabilistic translation rule to determine that a particular chromosome is good or bad. Now let us think how this algorithm should work to implement evolutionary computation. The algorithm in a flow chart expressed here randomly initializes the population which consists of a set of chromosomes which have a perspective to get a particular best results. Then we select the first individual, we evaluate its fitness. If it is fit then we perform a particular crossover and mutation and finally select the best chromosome which has a good objective or fitness function. If it is not fit then we go to select the next chromosomes. The operations used here are mechanisms like copying and swapping partial strings. The basic operators used are reproduction, simple crossover, mutation. Every chromosome is represented as binary strings. There is recombination of them at end point or at uniform. Mutation takes place bit wise, bit flipping with fixed probability. A parent selection is done from the fitness. Based on fitness we select the best of the population which is selected and there is a survivor selection and all children that survive replace the parents which the children have a better objective function. Speciality is the emphasis is given on crossover. Reproduction is the particular report that we are going to establish which selects individual proportion to its fitness. This operator copies to the next generation. Thus there is a natural selection and survival of the fittest of the best chromosomes. For a simple genetic algorithm representation a phenotype is encoded and this is produced as binary strings in the genome space. The simple genetic algorithm reproduction cycle is performed by the steps. First we select parents from the mating pool which are a set of chromosomes and this mating pool has a size called as a population size. We shuffle the mating pool. For each consecutive pair apply crossover with the probability PC. Otherwise the copy the parents directly. For each offspring apply mutation which has a probability PM which is a bit flip and independent for each bit. Replace the whole population with the resulting of springs. A single crossover is demonstrated here where we choose a random point on the two parents. We split the parents at its crossover point which is indicated by the vertical line and we create children by exchanging the tails and the probability of crossover is ideally taken between the range 0.6 to 0.9. Simple genetic algorithm second operator third operator is mutation. We alter each gene independently with the probability PM. PM is the probability of mutation called as a mutation rate. It is typically one upon population size and one upon chromosome length. At that particular point we exchange the bits and produce a child which might have more attraction for us to have a better objective function and a perspective solution developed. For selection we will be using the Rowlett wheel. The main idea is better individuals get higher chances. The chance is related to a fitness and using this Rowlett wheel technique where we have the probabilities expressed on the Rowlett wheel. The die being tossed on the Rowlett wheel and the selection being done to assign each individual a part of the Rowlett wheel and we spin the wheel n times to select n individuals for our basic population. For example here we have the probability of a to be 3, b to be 1 and c to be 2. This is the fitness value that a particular chromosome will have to have associated with it. The crossover function replaces randomly selected subtree of the individual with the randomly chosen subtree of another individual. Cross over point is also selected at random. We may have an endpoint crossover wherein at end points we are exchanging the tails of the particular chromosomes. The tails between the parents exchanged and then copied as children. We choose n randomly cross points split along these points, blue parts alternating between the parents and generalize of a single point still some positional bias is established wherever required. For a uniform crossover we will have a crossover going on in a uniform way and we assign heads to one parent, tails to the other, flip a coin for each gene of the first child, make an inverse copy of the gene for the second child, alternate this till all the genes are visited and inheritance is dependent on the position. For mutation we may change a single bit or a number of bits as required. Random alteration of the value of the string positions are done. It uses a flip function to change the time to a false or a vice versa and thus we are achieving better chromosomes. Here we remark that genetic algorithms are instances of random processes at work. A randomization established at the selection, a randomization established with the probability based orientation in the crossover and mutation, a randomization that is based using a Raoult selection of the particular individual to take part in the particular mating, a randomization where we are going to go for best objective optimized function to be established in a particular chromosome and mating and crossing are done also at random selection to produce more fit members. This concludes that genetic algorithm is an optimization process and it can be used for our particular references. To establish a reference for this particular content I have used the following. Thank you.