Open Source Algorithms - Page 5

Coach is a python framework that models the interaction between an agent and an environment in a modular way. With Coach, it is possible to model an agent by combining various building blocks, and training the agent on multiple environments. The available environments allow testing the agent in different fields such as robotics, autonomous driving, games and more. It exposes a set of easy-to-use APIs for experimenting with new RL algorithms and allows simple integration of new environments to solve. Coach collects statistics from the training process and supports advanced visualization techniques for debugging the agent being trained. Coach supports many state-of-the-art reinforcement learning algorithms, which are separated into three main classes - value optimization, policy optimization, and imitation learning. Coach supports a large number of environments which can be solved using reinforcement learning.

Conrad is both a high performance Conditional Random Field engine which can be applied to a variety of machine learning problems and a specific set of models for gene prediction using semi-Markov CRFs.

Esta é a documentação de um projeto de conclusão de curso feito na conclusão do curso de Engenharia de Computação-UNIVALI.É um conjunto de arquivos C e Matlab contendo a implementação da Lógica Fuzzy e simulações em sistema de enegia.

CuberGA project is the flexible Genetic Algorithms framework. Realize your ideas easy with deliveries of this project. Keywords: genetic algorithms, framework, permutation, mutation, crossover, genotype, selection, survival, Левченко Илья.

Cutting Problem solved by Genetic algorithms. The goal is to cut a rectangular plate of material into more smaller rectangles. The cuts must be rectangular and guillotinable.

This chess program changes its strength to give the best match against you. Eventually it learns to beat you specifically through learning alogirthms. Features included transposition tables and a elementary 3-piece endgame tablebase.

Detectron2 is Facebook AI Research's next generation software system that implements state-of-the-art object detection algorithms. It is a ground-up rewrite of the previous version, Detectron, and it originates from maskrcnn-benchmark. It is powered by the PyTorch deep learning framework. Includes more features such as panoptic segmentation, Densepose, Cascade R-CNN, rotated bounding boxes, PointRend, DeepLab, etc. Can be used as a library to support different projects on top of it. We'll open source more research projects in this way. It trains much faster. Models can be exported to TorchScript format or Caffe2 format for deployment. With a new, more modular design, Detectron2 is flexible and extensible, and able to provide fast training on single or multiple GPU servers. Detectron2 includes high-quality implementations of state-of-the-art object detection.

This project is an extended implementation of Knuth's "Dancing Links" algorithm and some use cases (e.g. Sudoku).

DisChoco is a Distributed Constraint Reasoning Solver. Several Distributed Constraint Reasoning algorithms are implemented (like ABT, AFC, Adopt, ...). Users may easily implement and test their algorithms with DisChoco. contact: wahbi[at]users[dot]sourceforge[dot]net

Dopamine is a research framework for fast prototyping of reinforcement learning algorithms. It aims to fill the need for a small, easily grokked codebase in which users can freely experiment with wild ideas (speculative research). This first version focuses on supporting the state-of-the-art, single-GPU Rainbow agent (Hessel et al., 2018) applied to Atari 2600 game-playing (Bellemare et al., 2013). Specifically, our Rainbow agent implements the three components identified as most important by Hessel et al., n-step Bellman updates, prioritized experience replay, and distributional reinforcement learning. For completeness, we also provide an implementation of DQN (Mnih et al., 2015). For additional details, please see our documentation. We provide a set of Colaboratory notebooks which demonstrate how to use Dopamine. We provide a website which displays the learning curves for all the provided agents, on all the games.

A teachable/trainable artificially intelligent music bot fundamentally inspired by how the new wave band Duran Duran composes music. This program utilizes many algorithmic/AI techniques/processes, including machine learning; which allow you to teach/train it to compose music which you prefer... and the technique which is the foundation of the design of DuranDuranbot, which was directly inspired by how Duran Duran writes music........ Called, "bit by bit circular composition"....... and it's explanation can be found here - https://scsynth.org/t/bit-by-bit-circular-composition/1107 This program is written in the SuperCollider programming language - https://en.wikipedia.org/wiki/SuperCollider Contact - ken_brant@ymail.com

Classical genetic algorithm suffers heavy pressure of fitness evaluation for time-consuming optimization problems. To address this problem, we present an efficient genetic algorithm by the combination with clustering methods. The high efficiency of the proposed method results from the fitness estimation and the schema discovery of partial individuals in current population and. Specifically, the clustering method used in this paper is affinity propagation. The numerical experiments demonstrate that the proposed method performs promisingly for well-known benchmark problems in the term of optimization accuracy.

EZvolve Foundation Classes is a set of data types and utility classes for use with evolutionary algorithms. Currently implemented support for bit string encoding, populations, fitnesses, fitness functions, probabilities, and probability vectors.

EasyGenetic is a highly efficient C++ framework for genetic master-slave algorithms. It heavily makes use of template metaprogramming to provide a fully customizable environment that accommodates most of user's needs without sacrificing efficiency.

The Free Connectionist Q-learning Java Framework is an library for developing learning systems. Keywords: qlearning, artificial intelligence, alife, neural nets, neural networks, machine learning, reinforcement learning unsupervised learning agents lejos

Emergent Mercenary is a 3D FPS with strategy overtones, and an excerise in emergent gaming. It should be easy to setup a scenario (e.g. hunt criminal) just by placing characters with certain AI into the game and see them interact.

Evokit is an integrated environment for developing evolutionary computation algorithms. It is built around Java and the Eclipse platform, and will support several different libraries including ECJ.

Flexible job shop scheduling problem (FJSP) is very important in many fields such as production management, resource allocation and combinatorial optimization. In the real manufacturing systems, each operation could be processed on more than one machine and each machine can also process several operations. This feature is known as flexibility. You can define your problem in this software and get an optimal solution as a Gantt Chart. This software is based on my M.Sc. thesis of Shahed university (Tehran, Iran).

Fuzzy numbers and basic fuzzy arithmetics (+, -, *, /, 1/x) implementation written in Java. Should be used as a library, but it also has a GUI which demonstrates its capabilities (using SWT widgets).

A genetic algorithm framework to allow the evolution of synapse weights and topologies of artificial neural networks.

This project uses massively parallel Graphics Processing Units(GPU) for neural network(Backpropagation) purposes.

Game Toolbox is a C# assembly designed to facilitate the creation of games and game prototypes. It contains no graphics code, does not depend on XNA, and is Mono-compatible. It provides implementations of A*, GOAP, a genetic algorithm, and more.

Gamoliyas is an open source John Conway's Game of Life game totally written in DHTML (JavaScript, CSS and HTML). Uses mouse and keyboard. Very configurable. This cross-platform and cross-browser game was tested under BeOS, Linux, *BSD, Windows and other

GeneThello (read jə-ˈne-ˈthe-lō), is an acronym for genetic othello, an othello (reversi) playing program which based on Genetic Algorithm (GA). In principle GeneThello consist of an othello program and a genetic algorithm system.

This project is a genetic algorithm engine able to be reused for other projects with minimal additional programming. The genetic algorithm engine currently plays many blackjack hands for the fitness function and produces a result similar to blackjack basic strategy. To see it in action, download the zip file and run either: GABlackjack_Demo.exe     (quick)   or GABlackjack_Long.exe       (slow, but it achieves better results). The code was written in C++, using MS Visual Studio 6.0 and MS Visual Source Safe 6.0. The genetic algorithm engine supports various mutation rates, ranked parental selection, stochastic sampling parental selection, cyclic crossover, crossover at each gene, cloning the best individual each generation, and creating random individuals each generation. To use the genetic algorithm engine to search for a different problem's solution, one needs to program a fitness function, the project settings, and a few virtual functions.