Do I need classification or regression to predict the availability of a user given some features?

Question

While studying data mining methods I have come to understand that there are two main categories:

• Predictive methods:

  • Classification
  • Regression

• Descriptive methods:

  • Clustering
  • Association rules

Since I want to predict the user availability (output) based on location, activity, battery level (input for the training model), I think it's obvious that I would choose "Predictive methods", but now I can't seem to choose between classification and regression. From what I understand this far, classification can solve my problem, because the output is "available" or "not available".

Can classification provide me with the probability (or likelihood) of the user being available or not available?

As in the output wouldn't just be 0 (not available) or 1 (for available), but it's be something like:

• $80\%$ available
• $20\%$ not available

Can this problem also be solved using regression?

I get that regression is used for continuous output (not just 0 or 1 outputs), but can't the output be the continuous value of the user availability (like the output being $80$ meaning user is $80\%$ available, implicitly the user is $20\%$ unavailable).

Answer 1

1. Yes. For instance, the popular softmax regression gives you probability distribution for each class.
2. Yes. Softmax is a regression over a set of discrete classes.

We can use regression for classification, the most common strategy is to grab the most likely class for the prediction.

Answer 2

Yes you can user either classification or regression according to your output requirement,

If you want labeled output, like either available or not available then classification should be used.

If you want the output in the form of % of availability then regression should be used.

Answer 3

You can use naive bayes classification and calculate posterior probabilities using prior beliefs or logistic regression can be used with sigmoid function.