Estimating Confidence Values of Individual Predictions by their Typicalness and Reliability

Although machine learning algorithms have been successfully used in many problems, and are emerging as valuable data analysis tools, their serious practical use is affected by the fact that often they cannot produce reliable and unbiased assessments of their predictions’ quality. There exist several approaches for estimating reliability or confidence for individual classifications, and many of them build upon the algorithmic theory of randomness, such as transduction-based confidence estimation, typicalness-based confidence estimation, and transductive reliability estimation. Unfortunately, they all have weaknesses: either they are tightly bound with particular learning algorithms, or the interpretation of reliability estimations is not always consistent with statistical confidence levels. In the paper we propose a joint approach that compensates the mentioned weaknesses by integrating typicalness-based confidence estimation and transductive reliability estimation into a joint confidence machine.