Title of the lecture
“New information fusion techniques and applications on image processing, classification and the computational brain”
In this talk we are going to focus on the problem of information fusion. More specifically, we are going to discuss some recent techniques which are based on the notion of aggregation function. We are going to relax the conditions required of these functions, specially those related to monotonicity, so we will introduce the class of pre-aggregation functions. We will show how we can use these new functions in several different problems, including edge detection or classification, for instance. In particular, we will apply them to a specific instance of neurocomputing: how to identify, from the electrical signals of the neural activity and in real time, whether a given subject is thinking of moving the left hand or the right hand. Considering this setting as binary classification problem, we will see that a specific class of pre-aggregation functions, built in terms of the so-called Sugeno integral, opens a very promising way to be explored in this problem.