Understanding about KERAS

Hi, everyone! I really believe to express my knowledge in pure and simple words!Today we are going to understand about one of the leading DEEP LEARNING library KERAS. We have already learnt about the deep learning concept, machine learning as well, if you did not read, read it first!

Introduction:

In our daily life, library is greatest way to express knowledge as a silence place, when we see it, we observe that it is source of information, collection of book and in a book a lot of information bound. But do you know what is library in computer programming language domain? It is the collection of classes every class has unique features and according to the need we use it. For example Math class library in C# give us a lot of math function without any code.

In python there are many useful library in deep learning concept but Keras in one of them which is one of the most popular library. So, in this Article we will discuss about KERAS!

What is Keras?

Keras is a Python library for deep learning that can run on top of Theanoor TensorFlow. Another simple definition, Keras is an open source neural network library which is written in Python.

For Deep learning concept we use Keras (Python Library) which run on Theano or TensorFlow. We read and understand the concept of deep learning in DEEP LEARNING post, in term of work this library play a vital role. This was made for fast and easy research and development in deep learning concept. It plays on PYTHON 2 or 3, can seamlessly execute on GPUS and CPUs.

Keras contains numerous implementations of commonly used neural network building blocks such as layers, objectives, activation functions, optimizers, and a host of tools to make working with image and text data easier. The code is hosted on GitHub, and community support forums include the GitHub issues page, and a Slack channel.

Several people thing about the concept of Keras is reserved for PYTHON only, but it’s allow user to productive deep models on smartphones (iOS and Android) also, and on we also on Java Virtual Machine.

François Chollet developed Keras, now let's understand the four principles of guiding of Keras.

Principles

Modularity: A model can be understood as a sequence or a graph alone. 
All the concerns of a deep learning model are discrete components that can 
be combined in arbitrary ways.
Minimalism: The library provides just enough to achieve an outcome, no frills 
and maximizing readability.
Extensibility: New components are intentionally easy to add and use within
 the framework, intended for researchers to trial and explore new ideas.
Python: No separate model files with custom file formats. Everything is 
native Python.

Resistance:

According to the latest report 200,000 user on November 2017 move towards Keras, and it was the 10th most citied tool in the KD Nuggets in 2018 software poll and registered a 22% usage.

It also allows use of distributed training of deep learning models on clusters of Graphics Processing Units (GPU).

User Experience.

Large adoption in the industry and research community.

Multi-backend, multi-platform.

Easy productization of models.

How to Install Keras

It is simple and straightforward to install, it is good if you did any work on PYTHON and SciPy before. In your machine setup the THEANO or TENSORFLOW  should be already installed. 

In this installation step we cover both platforms THEANO and TENSORFLOW.
By using PyPi th installation process are very easy;

sudo pip install keras
Recent Version of keras 1.0.0 at the time of writing you wil get, on command line you can also check the version of keras by using command. 
python -c "import keras; print keras -- version --"

the result of above command will be
1.0.0

So, can we upgrade the installation of keras, YES you can by using command 
sudo pip install -upgrade keras

In the upcoming Article we will learn KERAS by using in practical

 project using DISEASE IDENTIFICATION.

 For more update keep in touch!!!