I'm working building a genetic algorithm that will learn to play snake. I've worked out how to add/remove layers and neurons in the model, allowing the model's size to change through mutation. But this brings the problem of how to perform crossover when models will have different numbers/sizes of layers. If it was just the size of each layer that could change, it would be one thing. I can handle that. But I just can't figure out how to do …

My understanding is that these sorts of algorithms can evolve/mutate data to hone in on specific desirable areas in large/difficult to search parameter spaces. Assuming one does this successfully, how does one generate new data/sample from that desirable range? Does doing so utilize the same algorithm/structure, or would additional methods need to be introduced? If genetic algorithms improve the efficiency of a search by finding promising regions, can they be considered as a type of generative models? Or would that …

I can't get my deep genetic algorithm snake game to work and I can't figure out why. At this point, I think it must be either the crossover_rate/mutation_rate or the actual crossover code itself is wrong. I'm hoping someone here can help me with figuring out which. So, here's my understanding of deep genetic algorithms: You have a pool of agents. They're randomly generated. You have each of them run, tracking their fitness up until they die. When all agents …

I have hand writed classifiers (there are a lot of them). It's implemented as collection of rule sets IIF - THEN. I want to optimize the % of errors. There some classifiers witch have vey big % of False Positive and False Negative results. During my reserch about this problem i've found RIPPER alghorytm witch, seems like, was designed to solve this kind of problems. Also there are Multi Naive Bias alghorythm that can be helpfull. As far as i …

I have what I propose as a solution to my problem, however I haven't ever seen it mentioned in this way, so I worry that there is a valid reason not to do things this way. I have a dataset of > 100,000 events, where each event has a winner. I have plenty of data points, some data on the event itself, and some data on each entrant. The number of entrants in each event is variable, and I want …

There are two closely related techniques in genetic programming. One of them is grammar based genetic programming (GBGP), which uses context free grammar to generate a derivative tree which represents the program. The other is grammatical evolution which uses something called genomes and codons, which is then mapped to a phenotype, a realization. The part where I get confused is, the phenotype realization can also be represented as a derivative tree. The codon to rule mapping is done through a …

I have a feature set from which I want to select various combinations and permutations of the features. The length of a solution feature vector can range between , say 5 - 20 features , and the ordering of the features are important , meaning that feature vector ABC is different from BCA i.e they are sequential and depends on each others output. The goal is to find many near optimal solutions around optimal solutions and the solution space is …

I'm new to genetic algorithms, but I haven't found specific info about real-coded GA's. I want to do antenna array optimization by using the real values of antenna position, phase, and amplitude, but I've only found binary encodings. Although the binary encodings seem easier during crossover, but I would like to use real continuous values, but I don't know how to apply crossover. In MATLAB, I've been using an error value as a fitness function obtaining a vector with all …

I have a genetic algorithm for a vehicle routing problem with time windows and I need to implement certain modifications. I am not sure what would be the best chromosome representations. I have tasks which can be divided into 3 sub-tasks with certain ordered time windows, they have to processed in order and all 3 (they represent collecting certain goods in a storage, delivering them and returning packaging to another storage). In the algorithm crossover part these tasks are combined …

I have a keras model that is being trained using a genetic algorithm. This means that the performance of the model needs to be tested many times. To test the model performance the model is used as a control system in a simulation. In this simulation the model is evaluated in a tight loop so the speed of the forward pass of the model is very important. I am using a very small model. I used to be a dense …

how to apply the genetic algorithm as a feature selection for text classification in python I need to use GA to select most relevant feature in text classification

I'm trying to make a bot to the famous "Icy Tower" game. I rebuilt the game using pygame and I'm trying to build the bot using Python-NEAT. Every generation a population of 70 characters tries to jump to the next platform and increase their fitness. right now the fitness is the number of platforms they jumped on, each platform gives +10. The problem I'm facing is that the bot isn't learning good enough after 1000 generations the best score was …

Suppose you have a population of N individuals with fitness 1, 2, . . . , N (i.e., all individuals have a unique fitness value). Suppose you repeatedly apply tournament selection without replacement with tournament size s = 2 to this population, without doing crossover, mutation, and replacement. In other words, you run a genetic algorithm with selection alone. After a certain number of generations you will end up with a population consisting of N copies of the same individual. …

The exact problem is a crate of industrial parts, made by injection molding in very high quantities. The objective is to put as much parts as possible in one crate. This is done by a small robotic arm that take the part from the injection molding machine, cold it a little and put it in the crate. The shape of the parts can be a bit complex. It can be basically anything that can be molded into a 2 parts …

I am working on a problem set in which I have "Cost Per Click" and "Day of the week" as independent variables and "Profit" as dependent variable. I want to use this dataset to train a model to predict the optimal "Cost Per Click" for the next day. I have trained a random forest model on the training data. Now in order to predict the optimal cost per click for the next day, I am planning to use the "Day …

for a part of my project we have to create request packets for dns, ssdp and ntp protocols. My professor has said to use genetic algorithms to come up with the field values for the request packet the response to which is the biggest. I can see how we could achieve this for dns since we have a couple different fields for dns like qtype and qname among others but for ssdp requests, I don't think there are many fields? …

I'm looking for articles that use neural network as genetic algorithms and solving optimisation problems by it. Any reference are welcome, books, lectures, YouTube, etc. Basically I'm looking for general information in this field, everything can be helpfully advantages and disadvantages on using NN this way, recommendations, etc.

I am using a genetic algorithm to maximize a few hundred thousand real-valued variables. Each of the variables, $x_i$, has its own independent boundary condition. The fitness function uses each of these variables to compute another value and returns the sum of everything: $$fitness = g(x_1) + g(x_2) + g(x_3) \ + \ ...$$ This is taking incredibly long in python. In this situation, what do I gain by maximizing all values at the same time, i.e. using the genetic …

I have heard of genetic algorithms, but I have never seen practical examples and I've never got a systematic introduction to them. I am now looking for a textbook which introduces genetic algorithms in detail and gives practical examples how they are used, what their strengths are compared to other solution methods and what their weaknesses are. Is there any standard textbook for this?

What is a genetic algorithm, and what are its practical advantages over other algorithms? Is it similar to any commonly used machine learning algorithm like linear/logistic regression, neural networks, or tree-based methods like gradient boosting and random forests? I've heard it's based on "mutated" combinations of other models. Does this make it more like an ensemble?