The simplest form of regression: linear regression
It happens that, over time, scientists obtain a series of successive data. It is normal for them to be interested in discovering a trend in these data. As a simple example, imagine a lab that investigates the speed of growth of an herb, and keeps daily track of its size. They obtain a series of measurements not necessarily proportional to the number of days. This can be due to factors like temperature and humidity for example. But it is interesting to know that there is a mathematical way to discover if there is a trend, and even to give this trend a numerical value. A primary method to obtain this value is called "linear regression". There is a nice exposé of it on this Wikipedia page. In the article we find the image depicted in fig 1.
Fig. 1 An example of linear regression