As a hill-climbing algorithm, wouldn't this approach tend to result in the local optimum rather than the most general truths?
As a hill-climbing algorithm, wouldn't this approach tend to result in the local optimum rather than the most general truths?