machine learning

How can we design more effective stimulation strategies?

Rather than predicting perceptual distortions, one needs to solve the inverse problem: What is the best stimulus to generate a desired visual percept?