Could you re-formulate your question with more details? It is hard to give a good answer as long as you stay pure on that formula.

It is rather difficult to tell unless we know what x and y are, or what w1 and w2 signify. Can you clarify a little? Plus, what is being trusted, by whom, for what purpose? These are all important aspects of the problem.

Multilayer Perceptron or Classic classifiers like SVM and Random Forest should be able to provide a good estimate. However, this requires training data which in this case would be tuples of (trust,x,y)