TensorFlow is a Python library for fast numerical computing created and released by Google. It is a foundation library that can be used to create Deep Learning models directly or by using wrapper libraries that simplify the process built on top of TensorFlow. (www.TensorFlow.org)

University of Minnesota - Twin Cities

Python Libraries are chosen mostly based on their scalability and the area of tasks the program needs to specialize in. Here are some of common libraries that are in use:

- Keras - one of the most popular and open-source neural network libraries for Python

- PyTorch - Developed by Facebook, support for C++ with its C++ interface

- Scikit-learn - easy integration with different ML programming libraries like NumPy and Pandas

- NLTK(Natural Language Toolkit) - Python library for working with natural language processing

- Spark MLlib - Developed by Apache, enables easy scaling of your computations

- Theano - a robust library for carrying out scientific calculations on a large-scale

-TensorFlow - An open source library developed by Google for training neural networks.

Python Libraries for Machine Learning

Scikit-learn is an open-source machine learning library for programming languages. It provides support for  Support vector machine, random forest, gradient boosting, K-means and DBSCAN and classifications, regression and clustering algorithms, and design interoperable with numerical and scientific libraries(Numpy and Scipy). Like this, because Scikit-learn uses the scikit stack outside of the library, the biggest advantage is its compatibility with other libraries. Since the library has a unified interface, it is very simple to apply different methods.

Scikit-learn

NumPy is a Python library for scientific computing that contains support for some fundamental data structures used by scikit-learn, such as multi-dimensional arrays. Example import: `import numpy as np`

NumPy

Pandas is a Python library for data manipulation and analysis.
Example import: `import pandas as pd`

Pandas library in Python

BERT (Bidirectional Encoder Representations from Transformers) is a recent paper published by researchers at Google AI Language. It has caused a stir in the Machine Learning community by presenting state-of-the-art results in a wide variety of NLP tasks, including Question Answering (SQuAD v1.1), Natural Language Inference (MNLI), and others.


BERT’s key technical innovation is applying the bidirectional training of Transformer, a popular attention model, to language modelling. This is in contrast to previous efforts which looked at a text sequence either from left to right or combined left-to-right and right-to-left training. The paper’s results show that a language model which is bidirectionally trained can have a deeper sense of language context and flow than single-direction language models. In the paper, the researchers detail a novel technique named Masked LM (MLM) which allows bidirectional training in models in which it was previously impossible.

BERT Explained: State of the art language model for NLP

Natural Language Processing is a process of manipulating or understanding speech or text with software or machine. NLTK contains libraries for statistical language processing. There are many NLP libraries out there for use but NLTK is by far the most diverse. Other NLP libraries are best suited for specific tasks.

NLTK

PyTorch is an open source machine learning library. Benefits of using this library is that it offers easy to use APIs, is integrated with python well (meaning that code is readable, hence easier to debug), and offers dynamic computational graphs.

Pytorch

Theano is a python library that is built on top of NumPy. It has the ability to optimize, define, and efficiently evaluate mathematical expressions involving multi-dimensional arrays.

To install Theano, go to the terminal and type in  pip install Theano

Learn Before

Related