Machine learning has recently come to the fore in the past decade at an exponential rate. Large companies have now begun to pour significant resources into generating new and iterating on existing machine learning (ML)/deep learning (DL) libraries make taking advantage of power of ML much easier than it has been in the past. Good libraries provide:

  • Libraries that abstract away the greatly complex elements
  • Documentation and active backing my a competent team
  • Make it easy for a beginner can get up and started

Libraries like Sci-kit Lean and Tensorflow(ran by Google) are well covered across a variety of videos and third-party teams. Machine learning uses tested mathematical models generated by third party teams and used on a variety of problems like:

  • building intelligent systems
  • transcribing speech
  • detecting specific objects within a camera view
  • predict future stock prices


  • Tensorflow
  • Theano
  • Pytorch
  • Scikit-learn
  • Keras


  • Bonus libraries
    • MLflow
    • spaCy
      • PROS
        • The most popular and complete NLP library
        • Many 3rd party extensions
        • Fast sentence tokenization
        • Supports the largest number of languages compared to other libraries
      • CONS
        • Steep learning curve
        • Near the slowest of the bigger players
        • No integrated word vectors
        • No stock NN models