10 entry-level machine learning open source projects for programmers

Today we are going to introduce 10 open source machine learning projects suitable for getting started with AI programmers. To start contributing to an open source project, there are some prerequisites:

1. Learn a programming language: Since in open source contributions you need to write code to participate in development, you need to learn any programming language. Learning another language at a later stage is easy, depending on the needs of the project.

2. Familiarity with version control systems: These software tools help keep all changes in one place so that they can be recalled at a later stage when needed. Basically, they keep track of every modification made in the source code over time. Some popular version control systems are Git, Mercurial, CVS, etc. Among them, Git is the most popular and widely used in the industry.

1. Caliban

This is a machine learning project by tech giant Google. It is used to develop machine learning research workflows and notebooks in an isolated and reproducible computing environment. It solves a big problem. When developers are building data science projects, many times it is difficult to build a test environment that can demonstrate the project in real life. Therefore, Caliban is a potential solution to this problem.

10 entry-level machine learning open source projects for programmers

Caliban makes it easy to develop any ML model locally, run the code on the machine, and then try the exact same code in the cloud to execute on a larger machine. As a result, Dockerized research workflows are simple both on-premises and in the cloud.

2. Cornia

Kornia is a computer vision library for PyTorch. It is used to solve some general computer vision problems. Kornia is built on PyTorch and relies on its efficiency and CPU power to compute complex functions.

10 entry-level machine learning open source projects for programmers

Kornia is a set of libraries for training neural network models and performing image transformations, image filtering, edge detection, epipolar geometry, depth estimation, and more.

3. Analytics Zoo

Analytics Zoo is a unified data analytics and artificial intelligence platform that brings together TensorFlow, Keras, PyTorch, Spark, Flink, and Ray programs into one integrated pipeline. This can efficiently scale from laptops to large clusters to handle big data production. This project is maintained by Intel-analytics.

Analytics Zoo helps AI solutions by:

  • Helps easily prototype AI models.
  • 缩放得到有效管理。
  • 有助于将自动化流程添加到您的 ML 管道中,例如特征工程、模型选择等。

4. MLJAR 人类自动化机器学习

Mljar 是一个创建原型模型和部署服务的平台。 为了找到最佳模型,Mljar 搜索不同的算法并执行超参数调整。它通过在云中运行所有计算并最终创建集成模型来提供有趣的快速结果。 然后它会从 AutoML 培训中构建一份报告。 这不是很酷吗?

10 entry-level machine learning open source projects for programmers

Mljar 有效地训练用于二元分类、多类分类、回归的模型。

它提供两种接口:

  • 它可以在您的网络浏览器上运行 ML 模型
  • 在 Mljar API 上提供 Python 包装器。

10 entry-level machine learning open source projects for programmers

从 Mljar 收到的报告包含表格,其中包含有关每个模型分数和训练每个模型所需时间的信息。 性能显示为散点图和箱线图,因此很容易直观地检查哪些算法在所有算法中表现最佳。

5.DeepDetect

DeepDetect 是一个用 C++ 编写的机器学习 API 和服务器。如果想使用最先进的机器学习算法并希望将它们集成到现有应用程序中,那么 DeepDetect 很适合你。

10 entry-level machine learning open source projects for programmers

DeepDetect 支持各种各样的任务,如分类、分割、回归、对象检测、自动编码器。它支持图像、时间序列、文本和更多类型数据的有监督和无监督深度学习。 但是 DeepDetect 依赖于外部机器学习库,例如:

  • 深度学习库:Tensorflow、Caffe2、Torch。
  • 梯度提升库:XGBoost。
  • 使用 T-SNE 进行聚类。

6. Dopamine

Dopamine 是科技巨头 Google 的一个开源项目。 它是用 Python 编写的。它是一个快速原型强化学习算法的研究框架。

Dopamine 的设计原则是:

  • 轻松实验:Dopamine 使新用户可以轻松运行实验。
  • 它紧凑而可靠。
  • 它还有助于结果的重现性。
  • 它很灵活,因此使新用户可以轻松尝试新的研究思路。

7. TensorFlow

Tensorflow 是 GitHub 上最著名、最受欢迎和最好用的机器学习开源项目之一。它是一个开源软件库,用于使用数据流图进行数值计算。它有一个非常易于使用的 Python 接口,并且没有其他语言中不需要的接口来构建和执行计算图。

10 entry-level machine learning open source projects for programmers

TensorFlow 提供稳定的 Python 和 C++ API。 Tensorflow 有一些惊人的用例,例如:

  • 在语音/声音识别中
  • 文本库应用程序
  • 图像识别
  • 视频检测
  • …还有很多!

提到图像识别与视频检测技术,不得不提目前在各个领域很火的AI+视频技术,将AI检测、智能识别技术融合到各个视频应用场景中,如:安防监控、视频中的人脸检测、人流量统计、危险行为(攀高、摔倒、推搡等)检测识别等。典型的示例如EasyCVR视频融合云服务,具有AI人脸识别、车牌识别、语音对讲、云台控制、声光告警、监控视频分析与数据汇总的能力。

10 entry-level machine learning open source projects for programmers

8.PredictionIO

它建立在最先进的开源堆栈之上。 该机器学习服务器专为数据科学家设计,可为任何 ML 任务创建预测引擎。 它的一些惊人功能是:

  • 它有助于在可定制的生产模板上快速构建和部署引擎作为 Web 服务。
  • 部署为 Web 服务后,即可实时响应动态查询。
  • 它支持机器学习和数据处理库,如 OpenNLP、Spark MLLib。
  • 它还简化了数据基础设施管理

10 entry-level machine learning open source projects for programmers

9.Scikit-learn

它是一个基于 Python 的免费软件机器学习工具库。它提供了用于分类、回归、聚类算法的各种算法,包括随机森林、梯度提升、DBSCAN。

10 entry-level machine learning open source projects for programmers

这是建立在必须预先安装的 SciPy 之上,以便可以使用 sci-kit learn。 它还提供以下模型:

  • 集成方法
  • 特征提取
  • 参数调优
  • 流形学习
  • 特征选择
  • 降维

注意:要学习 scikit-learn 遵循文档:
scikit-learn.org/stable/

10. Pylearn2

Pylearn2 is the most popular machine learning library among all Python developers. It is based on Theano. You can write its plugins using mathematical expressions, while Theano needs to be optimized and stabilized.

10 entry-level machine learning open source projects for programmers

It has some great features like:

  • The "default training algorithm" used to train the model itself
  • Model Estimation Criteria
  • Score match
  • cross entropy
  • log likelihood
  • Dataset Preprocessing
  • Contrast normalization
  • ZCA whitening
  • Patch extraction (for implementing convolution-like algorithms)

Guess you like

Origin juejin.im/post/7020316021399486495