mlpack VS Caffe

Does C++ have the equivalent of Pandas or Apache Spark? Are there extensive libraries that exist/are being developed that allow you to perform operations with data? Or do people just use a combination of Python & its various libraries (NumPy etc)? If we leave aside the data bit, are there libraries that allow you to develop ML models in C++ (mlpack for instance ) faster & more efficiently compared to their Python counterparts (scikit-learn)? On a more general note, how does C++ fit into the routine of a Quant Researcher? And at what scale does an organization decide they need to start switching to other languages and spend more time developing the code ?

mlpack is a great library for machine learning in C++. It's very fast and not too much of a learning curve.

This toolkit was originally part of the mlpack machine learning library (https://github.com/mlpack/mlpack) before it was split out into a separate, standalone effort.

Github Repository: https://github.com/mlpack/mlpack Developed By: Community, supported by Georgia Institute of technology Primary purpose: Multiple ML Models and Algorithms

Click to Learn more...

Caffe de BAIR

Caffe is a DL framework just like TensorFlow, PyTorch etc. OpenPose is a real-time person detection library, implemented in Caffe and c++. You can find the original paper here and the implementation here.