TensorFlow, why was python the chosen language?

Question

I recently started finding out deep learning & machine learning techniques, and I started sorting out frameworks that simplify the method of building a net & training it, then I found TensorFlow, having very little expertise within the field, for me, it looks that speed may be a huge issue for creating a big machine learning system even a lot of if operating with deep learning, thus why python was chosen by Google to form TensorFlow? Won’t it be efficient to use a language which can be compiled and not interpreted?

Lastly, mention some advantages Python has over C++ in machine learning.

Answer 1

The core of Tensorflow is not written in Python but it is written in a combination of C++ and CUDA.

Python is chosen by Google over other languages mainly because-

1. Python is easy to integrate and has control over C++ backend.

2. Most of the data scientists and machine learning experts are comfortable with Python.

3. Numpy makes it easy to do pre-processing in Python