Creation of a Target Variable So i'm new to machine learning and data science, i've been looking tutorials and i'm working on self made project at the moment and i'm having an analysis paralysis with the data preprocessing portion. I think i'm over complicating this but here are some questions I have.
So for example, I have survey data from people about a store and I hope to use this data to help them find out whether or not they're willing to spend money on the store.
I have things like median household income, age, gender, whether or not they're happy, but for what I want my target variable to be would be something along the lines of "Would this person buy stuff from this store," this obviously isn't in the data set. How would I go about creating this as my target variable? Is this feasible to do in the first place?
All tutorials online have their target variable already in the data set so they can train on it, would creating the target variable with 0 values be feasible? Would I need to create logic to fill the values from already established predictor variables? 
 A: Without a target variable, you cannot do supervised machine learning. After all, you don't know whether any of these people would shop at the store or not, so any prediction your model produces will be a complete guess, and you won't have any way to tell whether it's right or not.
If you just want to do a toy problem, you could build a synthetic target variable based on some input features like income or gender. But that relationship will be entirely contrived however you see fit, and your predictive model should just recover those relationships to arrive at a model that's pretty similar to however you generated the data in the first place. This can be useful to get experience in the mechanics of just running some methods, but you shouldn't base any domain inferences on your output, since you are the one that's defining true model you're attempting to learn. Toy problems such as this can be useful for benchmarking different methods, though, since you know exactly what the true underlying model is. You know the target variable is, in fact, predictable in the first place, and exactly which features it depends upon.
