We invite you to the seminar, which will be held on 5 September 2017 at 16.00 in ONU. Mechnikov.
”Algorithmic Learning: conformal predictions and testing”
The talk reviews a modern machine learning technique called Conformal Predictors (CP). The technique is based on algorithmic notion of randomness and allows us to make reliable predictions with valid measures of confidence for individual examples. This approach guarantees that the overall accuracy can be controlled by a required confidence level. Unlike many conventional techniques the approach does not make any additional assumption about the data beyond the iid assumption: the examples are independent and identically distributed. The way to test this assumption is also described.
The talk outlines the basic ideas of Conformal Predictors and then illustrates the technique with applications to several industrial problems. In particular, we outline an application of CP in pharmaceutical industry for drug discovery in ongoing European Union Horizon 2020 project.