On Threshold Correlation with Application to Studying the Relationship of Temperature and Relative Humidity

What is it about?

This study finds subset level correlation using thresholds from continuous weather variables humidity and temperature. between A continuous variable is a variable that can take on any value, for example temperature can be any value such as 100 degree. Continuous variables are important for all types of studies you can think of AI, Machine Learning, Economics, Life Science. These studies often focus on finding relation between continuous variables, such as if value of one variable increase, does other increase/decrease/stays usual. For example, when COVID rate (a continuous variable) increased, number of people going to hospital (another continuous variable) increased. Correlation finds these linear trend (does increase of one variable affect another variable ). We found out that, if you take subset of values from any existing study, correlation changes significantly. For example, within the same data, that shows high positive linear relation, some subset can show positive relation, while other can show negative relation. We proposed thresholds (values from variables) to separate the subset and study these relations using temperature and humidity variables.,

Featured Image

Why is it important?

Continuous variable are ubiquitous in any scientific study we can think of, such as AI, Machine Learning, Economics, Life Science, astronomy etc. Also correlation is a very popular method always been applied to find relation among continuous variables. Our approach shows another way of extracting knowledge using correlation and the subset level relation we extract can help to discover unseen patterns often ignored by the scientist.

Perspectives