Abstract:Tensor network diagram (graphical notation) is a useful tool that graphically represents multiplications between multiple tensors using nodes and edges. Using the graphical notation, complex multiplications between tensors can be described simply and intuitively, and it also helps to understand the essence of tensor products. In fact, most of matrix/tensor products including inner product, outer product, Hadamard product, Kronecker product, and Khatri-Rao product can be written in graphical notation. These matrix/tensor operations are essential building blocks for the use of matrix/tensor decompositions in signal processing and machine learning. The purpose of this lecture note is to learn the very basics of tensors and how to represent them in mathematical symbols and graphical notation. Many papers using tensors omit these detailed definitions and explanations, which can be difficult for the reader. I hope this note will be of help to such readers.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: **How to clearly and intuitively represent the basic concepts, operations and decomposition methods of tensors through graphical notation**.
Specifically, the goals of the paper include:
1. **Simplify the understanding of tensors**: Use graphical notation to simplify the understanding of tensors and their operations. Graphical notation can display complex tensor multiplications in a graphical way, making these operations more intuitive and easier to understand.
2. **Unify the representation method**: Provide a unified graphical representation method for various operations of matrices/tensors (such as inner product, outer product, Hadamard product, Kronecker product and Khatri - Rao product, etc.). This helps readers better understand the essence of these operations.
3. **Help beginners get started**: For beginners, understanding and mastering tensor decomposition may be a long process. By introducing graphical notation, the difficulty of learning tensor decomposition can be significantly reduced, enabling beginners to master relevant knowledge more quickly.
4. **Fill the literature gap**: Many papers using tensors omit detailed definitions and explanations, which causes difficulties for readers to understand. This paper aims to fill this gap, explain in detail the basic concepts of tensors and their graphical representation methods, and help readers better understand the content in relevant literature.
### Main content of the paper
- **Basic concepts of tensors**: Introduce the concepts of vectors, matrices and tensors, and explain the different modes of tensors.
- **Tensor operations**: Describe in detail the reshaping, calculations and decompositions of tensors, and illustrate them with graphical notation.
- **Application of graphical notation**: Show how to use graphical notation to represent various tensor operations, such as inner product, outer product, Hadamard product, Kronecker product and Khatri - Rao product, etc.
### Why is this problem important?
Tensors have wide applications in fields such as signal processing and machine learning, and tensor decomposition is one of the key techniques in these applications. However, it is not easy to understand and master tensors and their operations, especially for beginners. By introducing graphical notation, the understanding of these concepts can be greatly simplified, thus promoting the research and development in related fields.
For example, in deep learning, tensor networks have been applied to the optimization and acceleration of neural networks, and in quantum physics, tensor networks are also used to simulate complex quantum systems. Therefore, mastering the graphical representation methods of tensors and their operations is of great significance for researchers engaged in these fields.
### Summary
By introducing graphical notation, this paper aims to simplify the understanding of tensors and their operations, help beginners master relevant knowledge of tensor decomposition more quickly, and provide a more intuitive tool for researchers engaged in fields such as signal processing and machine learning.