Page 75 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

Chapter 3 Application, algorithm, tools directly related to deep learning 63

2. Tools used in deep learning

2.1 TensorFlow
TensorFlow is an open-source machine learning framework
for all programmers. It is used for implementing all machine
learning and deep learning applications. To develop and research
on fascinating ideas on artificial intelligence, Google team
purposely created TensorFlow. TensorFlow is designed in Python
programming language; hence it is considered easy to interpret
and use [1].
TensorFlow is a new software library or framework, designed
by the Google team to implement machine learning and deep
learning concepts in the easiest manner. It merges the computa-
tional algebra of optimization techniques for complex calculation
of many mathematical equations and expressions.
Let us now consider the following important features of
TensorFlow:
• It includes a feature that optimizes and calculates mathemat-
ical expressions very easily with the help of single- and multi-
dimensional arrays called tensors.
• It includes a programming and technical support of entire
deep neural networks and machine learning techniques.
• It includes high scalable characteristics of computation with
different data sets.
• TensorFlow uses GPU computing and automating manage-
ment. It also includes a distinct feature of optimization of
memory and the data used [1].

2.1.1 Tensor data structure
Tensors are used as the common data structures in Tensor-
Flow language. Tensors connecting all the edges in any flow
diagram are called the data flow graph. Tensors are defined as
multidimensional array or list [1].
Tensors are identified by three parameters, namely, rank,
shape, and type.

2.1.2 Rank
Unit of dimensionality for tensor is called rank. It identifies the
different dimensions of the tensor. A rank of a tensor can be
depicted as the order or n-dimensions of a tensor defined.