As you all know, in ML, if the Dependent feature is continuous in nature, then Regression models are applied. But, if Dependent feature is categorical in nature, then Classification algorithm are used.

As you can see in this graph (https://i.sstatic.net/9Q3wfudK.png), there are max. numbers of data points are repeated which suggests that they are forming categories.

Then, why here regression is being used?

Here is the dataset:(https://drive.google.com/file/d/1vTIiQ0NZKgBI-EfpGzfPKHx1VaAdEYdH/view?usp=sharing)

I discussed this question with my peers and teacher. They all are saying that regression is used for prediction, but no one can explain how they concluded that regression should be applied instead of classification.