I’m new to prediction techniques such as linear regression and time series analysis. Here’s my problem. My app is something like a todo-app. It has many projects. A project has many parts. A task has a priority and a number of tasks. Suppose there’s no constraint on the work. I want to predict the time to complete a work based on historical data for the same project. For linear regression, I have this:
y: The time to complete a workx1: Priority of a work (a category variable)x2: The number of tasks in a work.
While I know this model is not accurate enough, I still want to give a suggestion about the expected time because I cannot collect enough data to affect y. Is this okay? Can I use time series analysis instead?