Top Page | Upper Page | Contents | About This Site | JAPANESE

Analysis of Many vs Many

Multi-Regression Analysis is the method for one Y and many X. There are methods for many Y and many X.

The goal of this analysis is to understand the relationship in the simple way.

Classification by the Type of Matrix

I classify the approaches of many vs many analyasis by the type of matrix. There are three types, A-B type, A-A Typeand AB-C type.

A-B type can be changed into A-A type. AB-C type can be changed into A-B type. This change is one of the hint of Non-Developed Wide Area of Data Science .
A-B A-A AB-C

Graph of A-A

In the study of network , A-A type is often used. View of Network is used.

Multi Dimensional Scaling is used if relationship is the number of distance.

AHP

AHP uses the group of matrix. The main part of AHP uses A-B type data. Bu sub part uses A-A type data.

Many Applications of A-B

Statistics uses A-B type as "A is the variable. B is the sample." Many methods of Multi-Variable Analysis and Data Mining start A-B type data.

In Text Mining , A is the word. And B is the text. And Co-occurrence is studied by A-B type data.

Q-Analysis is also used for A-B type.




NEXT Matrix Decomposition