Tensors in Neural Networks--CNN 2

In Deep Learning, we store our data in the form of Tensors.

Tensors can also be defined as a generalization of matrices to an arbitrary number of dimensions.

Scalars (0D Tensors)

A tensor that contains only one number is called a scalar, a scalar-tensor has 0 axes (ndim == 0). The number of axes of a tensor is also called it’s rank(张量的轴数也叫做它的秩).

a = 0

Vectors (1D tensors)

An array of numbers is called a vector, or 1D tensor. A 1D tensor is said to have exactly one axis.

array([15, 2, 7, 4])

Matrices (2D tensors)

A matrix has two axes (often referred to as rows and columns).

>>> x = np.array([[5, 78, 2, 34, 0], [6, 79, 3, 35, 1], [7, 80, 4, 36, 2]])
>>> x.ndim
2

3D Tensor

An image tensor is always 3D : One dimension for its height , one for its width and a third one for its color channel.

4D Tensor

A 4D Tensor is an array of 3D Tensors.

5D Tensor

A 5D Tensor is an array of 4D Tensors, like so: