Python machine learning libraries, which ones have you used?

  As we all know, python has a lot of rich third-party libraries, which can provide more help for our work, which is one of the reasons why everyone learns python. So do you know what are the commonly used libraries for Python machine learning? How many have you used?

  1. Pipenv: It is a side project of Kenneth Reitz, whose main purpose is to integrate other software packages into python. The tool can generate a Pipfile.lock file, allowing you to build more deterministically and avoid BUG that is difficult to find.

  2. PyTorch: Facebook's deep learning framework, derived from the Torch framework, is based on the python language, and realizes the dynamic calculation graph paradigm. It has become one of the preferred frameworks for many researchers with strong scalability.

  3. Caffe2: Supports distributed training and deployment. The new CPU and CUDA-supported hardware are more suitable for research and large-scale deployment.

  4. Pendulum: It is a substitute for the python standard datetime. You can easily integrate with modern code and use its functions when you need it.

  5. Dash: A pure python open source library that can build web applications, especially data visualization web applications. It is built on Flask, Plotly and React, and provides abstract interfaces for these frameworks to achieve efficient development.

  6. PyFlux: Python open source library developed specifically for time series. Time series research is a subfield of statistics and economics, which can be used to describe the behavior of time series.

  7. Fire: An open source library that can automatically generate a command line interface for any python project.

  8. Imbalanced-learn: Provides related technologies to solve the problem of data imbalance. In addition, it is compatible with scikit-learn and is very useful.

  9. FlashText: It proves the importance of algorithm and data structure design. Even in the face of simple problems, better algorithms can easily surpass the simple implementation running on fast CPUs.

  10. Luminoth: An open source computer vision Python toolkit built for TensorFlow and Sonnet.


Guess you like

Origin blog.51cto.com/15052541/2678338