Page 76 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

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

2.1.3 Shape
Many numbers of rows and columns combined together
define the shape of Tensor.
2.1.4 Type
Type specifies the data type assigned to all Tensor's elements.
A user needs to consider the features for building a Tensor.
• Build an n-dimensional array using rows and columns.
• Convert the n-dimensional array.
TensorFlow includes various dimensions. The dimensions are
described as two types. The various dimensions of Tensorflow is
depicted in Fig. 3.3 [1].

2.1.5 One-dimensional Tensor
One-dimensional Tensor is a normal array structure, which in-
cludes one set of values of the same data type.

2.1.6 Two-dimensional Tensor
Sequences of arrays are used for creating “two-dimensional
Tensor.”
TensorFlow includes a visualization tool, which is called the
Tensor Board. It is used for analyzing sequential data flow graph
and also used to estimate machine learning models. The impor-
tant feature of Tensor Board includes distinct types of statistics
about the different parameters and details of graph in vertical
alignment.

1d-tensor 2d-tensor 3d-tensor

4d-tensor 5d-tensor
Figure 3.3 Various dimensions of TensorFlow.