- SkillEnable

# Introduction to Tensorflow

Updated: Sep 18, 2020

**What does the term Tensorflow mean? **

TensorFlow is an end to end platform for quick building and deploying of Machine Learning models. It is a complete ecosystem which efficiently enables solving of challenging and complicated world problems with machine learning.

It is a Python-friendly open source library which has many numerical computations. This makes machine learning much easier and faster.

Machine learning is a very complex discipline but Tensorflow, formulated by Google Brain makes it much simpler. It eases the process of training models,compiling data and categorizing future results.

**Unique features of Tensorflow**

Tensorflow deftly works with mathematical expressions involving huge numbers with multidimensional arrays.

It provides a well supported deep neural network with complex machine learning concepts which needs a lot of effort and time to analyse otherwise.

It provided both GPU and CPU computing where the same code can be executed on both architectures without any problem.

It has a good scalability of computation across machines and huge data sets. Therefore it eases the functions.

Another important feature of Tensorflow is its flexibility in operations. In other words it provides modularity. It also offers you the option standalone in the part you wish to add.

As it has been built by the Google Brain team, it has a large community. There is a huge team of software engineers working on its stability continuously.

One of the best things about the machine learning library provided by Tensorflow is that it has an open source. Therefore anyone with an internet connection can smoothly use it.

One of the unique features of Tensorflow is that it provides ‘feature columns’. These columns act as intermediaries between raw data and estimators, bridging input data with the clients model

TensorFlow lets software engineers and developers create graphs. These graphs explain the logic behind data moving through a sequence of processing nodes. Each node in the main graph represents a mathematical equation. It also creates a connection or edge between two nodes of the same graph in a multidimensional data tensor.

This unique high level programming language gives the above mentioned data the form of all Python languages. Python (blog- url) then provides an easy method to express high-level abstractions that can be put together. Nodes and tensors in TensorFlow are Python objects, whereas the TensorFlow applications are actually Python applications.

However the actual math equations are not performed in Python. It is created by the help of the large library provided by Tensorflow. This library helps in the transformations that are available through TensorFlow also written as high-performance C++ binaries. Thus the only work of Python is directing traffic between the pieces, and providing high-level programming abstractions so that the pieces can be together.

TensorFlow applications can be run on most of the devices, be it an iOS or an Android device. It supports CPUs as well as GPUs. TensorFlow can also be used on Google’s custom TensorFlow Processing Unit (TPU), if you use Google's own cloud. The latest version of Tensorflow is Tensorflow 2.0 which was released by Google in the year of 2019. It had a changed framework which was based on the user’s feedback. This version provides an easy method to work with Keras API.

**Steps to install Tensorflow **

conda create --name TensorflowEnv biopython

source activate TensorFlowEnv

pip install --upgrade pip

pip install tensorflow

Simple expression in Tensorflow

To equate functions like y = 5*x + 13 through Tensorflow programming, follow the below mentioned steps.

__Formula__

constant(value, dtype=None, shape=None, name='Const', verify_shape=False)

Therefore

z = tf.constant(5.2, name="x", dtype=tf.float32)

import tensorflow as tf

x = tf.constant(-2.0, name="x", dtype=tf.float32)

a = tf.constant(5.0, name="a", dtype=tf.float32)

b = tf.constant(13.0, name="b", dtype=tf.float32)

y = tf.Variable(tf.add(tf.multiply(a, x), b))

init = tf.global_variables_initializer()

with tf.Session() as session:

session.run(init)

print session.run(y)

The above mentioned coding expresses a simple equation in the Tensorflow fashion.

Tensorflow is an upcoming language with high level scalability and flexibility. It has a large library and feature columns.

To learn more about this programming language, visit (https://www.skillenableacademy.com/)

Stay tuned for more updates on industry relevant tools!

**References**

Infoworld