Open Source Python Algorithms - Page 3

BTK contains C++ and Python libraries that implement speech processing and microphone array techniques such as speech feature extraction, speech enhancement, speaker tracking, beamforming, dereverberation and echo cancellation algorithms. The Millennium ASR provides C++ and python libraries for automatic speech recognition. The Millennium ASR implements a weighted finite state transducer (WFST) decoder, training and adaptation methods. These toolkits are meant for facilitating research and development of automatic distant speech recognition.

DualPipe is a bidirectional pipeline parallelism algorithm open-sourced by DeepSeek, introduced in their DeepSeek-V3 technical framework. The main goal of DualPipe is to maximize overlap between computation and communication phases during distributed training, thus reducing idle GPU time (i.e. “pipeline bubbles”) and improving cluster efficiency. Traditional pipeline parallelism methods (e.g. 1F1B or staggered pipelining) leave gaps because forward and backward phases can’t fully overlap with communication. DualPipe addresses that by scheduling micro-batches from both ends of the pipeline in a bidirectional fashion—i.e. some micro-batches flow forward while others flow backward—so that computation on one partition can coincide with communication for another.

Evolutionary Algorithm is an educational Python project that demonstrates evolutionary computation techniques such as genetic algorithms, evolution strategies, and neuroevolution in a clear and accessible way. Rather than being a single monolithic library, this repository provides a series of self-contained examples showing how different population-based search methods solve optimization problems and adapt candidate solutions over generations. Users can explore basic genetic algorithm setups, match phrase examples, pathfinding challenges, and microbial GA variants, as well as evolution strategy approaches like NES. The project also links classical evolutionary approaches with neural networks, illustrating how evolution can be used for model training in reinforcement learning and supervised contexts.

FATE (Federated AI Technology Enabler) is the world's first industrial grade federated learning open source framework to enable enterprises and institutions to collaborate on data while protecting data security and privacy. It implements secure computation protocols based on homomorphic encryption and multi-party computation (MPC). Supporting various federated learning scenarios, FATE now provides a host of federated learning algorithms, including logistic regression, tree-based algorithms, deep learning and transfer learning. FATE became open-source in February 2019. FATE TSC was established to lead FATE open-source community, with members from major domestic cloud computing and financial service enterprises. FedAI is a community that helps businesses and organizations build AI models effectively and collaboratively, by using data in accordance with user privacy protection, data security, data confidentiality and government regulations.

FRODO is a Java platform to solve Distributed Constraint Satisfaction Problems (DisCSPs) and Optimization Problems (DCOPs). It provides implementations for a variety of algorithms, including DPOP (and its variants), ADOPT, SynchBB, DSA...

A cross-platform library that computes fast and accurate SIFT image features. libsiftfast provides Octave/Matlab scripts, a command line interface, and a python interface (siftfastpy). Optimized with SIMD instructions and OpenMP .

LZO is a compression library which is widely used around the world. The main problem with LZO is that it is absolutely not human readable. People have done crazy stuff to get LZO to run in their language. Usually it implies inline assembly or trying to execute data which actually contains machine code. This is sick. Whoever is responsible for this sorry situation ought to be ashamed. So I'm going to deobfuscate LZO and provide a ISC implementation of this algorithm in Python and C. In addition, I will provide a textual description of the algorithm so that it can be easily ported to any programming language. I expect a severe performance degradation, but I leave optimizing for speed to other people.

This is a python implementation that handles floating points correctly,there are still some bugs but I'm working on it . The point was to work around some stuff that made no sense for floats,like 0.1+0.2 == 0.3 is false.

Free and easy to implement algorithm that can be used for resource leveling in Gantt projects. Uses resources assignment policies, availibility calendars, tasks\' priorities, dependencies, sizings, complexity due to human communication.

C++, Matlab and Python library for Hidden-state Conditional Random Fields. Implements 3 algorithms: LDCRF, HCRF and CRF. For Windows and Linux, 32- and 64-bits. Optimized for multi-threading. Works with sparse or dense input features.

HDRFlow is a framework to process high-dynamic range (HDR) and RAW images. It's written in C++, and is both cross-platform and hardware accelerated on modern GPUs.

This program generates customizable hyper-surfaces (multi-dimensional input and output) and samples data from them to be used further as benchmark for response surface modeling tasks or optimization algorithms.

Institute of Technology, Blanchardstown Computer Science code by the class of 2007-2011 on course BN104. In this project we are open sourcing all of our project work to the public in the hopes it can be reused, built-upon, and used in education.

The Jsolution package is a collection of programs, scripts,libraries and documents that i have written and use in my daily life.

KaHaShEr is an application to check the integrity of a file using MD5, SHA1, SHA256 and SHA512 checksum. Designed to be smart and easy to use, it includes a feature that can detect a hexadecimal value in the clipboard and automatically perform the verification while notifying you with a very expressive sound of the outcome of this verification.

The project contains algorithms for locating the most central groups in complex networks. In particular implementations of various centrality measures and heuristic search algorithms. All functionality is exposed via XML-RPC for easier exploitation.

LASS : Library of Assembled Shared Source. Library of C++ code for scientific purposes.

The LisBON Framework is an adaptable framework for developing new parallel Memetic Algorithms (hybrid search algorithms for efficiently solving optimisation problems).

MADDPG (Multi-Agent Deep Deterministic Policy Gradient) is the official code release from OpenAI’s paper Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments. The repository implements a multi-agent reinforcement learning algorithm that extends DDPG to scenarios where multiple agents interact in shared environments. Each agent has its own policy, but training uses centralized critics conditioned on the observations and actions of all agents, enabling learning in cooperative, competitive, and mixed settings. The code is built on top of TensorFlow and integrates with the Multiagent Particle Environments (MPE) for benchmarking. Researchers can use it to reproduce the experiments presented in the paper, which demonstrate how agents learn behaviors such as coordination, competition, and communication. Although archived, MADDPG remains a widely cited baseline in multi-agent reinforcement learning research and has inspired further algorithmic developments.

MLPACK is a C++ machine learning library with emphasis on scalability, speed, and ease-of-use. Its aim is to make machine learning possible for novice users by means of a simple, consistent API, while simultaneously exploiting C++ language features to provide maximum performance and flexibility for expert users. * More info + downloads: https://mlpack.org * Git repo: https://github.com/mlpack/mlpack

Multi-categorization Recommendation Adjusting (MRA) is to optimize the results of recommendation based on traditional(basic) recommendation models, through introducing objective category information and taking use of the feature that users always get the habits of preferring certain categories. Besides this, there are two advantages of this improved model: 1) it can be easily applied to any kind of existing recommendation models. And 2) a controller is set in this improved model to provide controllable adjustment range, which thereby makes it possible to provide optional modes of recommendation aiming different kinds of users.

The market assumes an environment in which farmers sell their produce through brokers and traders locate produce to buy through brokers. The major aim of the simulator is to experiment with various reputation mechanisms to manage bottlenecks and to model various adversarial scenarios. The market is aimed to simulate agricultural trade in developing countries. It is written in python and mysql database on Linux.

Math tools in Python to tackle down problems in Operational Research fields. Comes with a Django based web interface to allow remote access to complex simulation means.

TiVo style recommendation engine for MythTV. MythMagic selects and automatically records shows from your program guide based on previous viewing habits. Recordings can be accepted or rejected to improve recommendation accuracy.

The NUMIPAD library implements several methods/algorithms to solve inverse problems and adaptive decomposition (i.e. Tikhonov regularization, Total Variation, Basis Pursuit, etc.)