Abstract:Bayesian averaging over classification models allows the uncertainty of classification outcomes to be evaluated, which is of crucial importance for making reliable decisions in applications such as financial in which risks have to be estimated. The uncertainty of classification is determined by a trade-off between the amount of data available for training, the diversity of a classifier ensemble and the required performance. The interpretability of classification models can also give useful information for experts responsible for making reliable classifications. For this reason Decision Trees (DTs) seem to be attractive classification models. The required diversity of the DT ensemble can be achieved by using the Bayesian model averaging all possible DTs. In practice, the Bayesian approach can be implemented on the base of a Markov Chain Monte Carlo (MCMC) technique of random sampling from the posterior distribution. For sampling large DTs, the MCMC method is extended by Reversible Jump technique which allows inducing DTs under given priors. For the case when the prior information on the DT size is unavailable, the sweeping technique defining the prior implicitly reveals a better performance. Within this Chapter we explore the classification uncertainty of the Bayesian MCMC techniques on some datasets from the StatLog Repository and real financial data. The classification uncertainty is compared within an Uncertainty Envelope technique dealing with the class posterior distribution and a given confidence probability. This technique provides realistic estimates of the classification uncertainty which can be easily interpreted in statistical terms with the aim of risk evaluation.

Comparison between Thermoelectric and Bolometric Microwave Power Standards

Comparison Among Coaxial Microcalorimeter Models

RF/MW power standard realization without unitary efficiency assumption at dc/LF

Improvement of primary power standard through international comparison feedback

Thermoelectric Sensors as Microcalorimeter Load

Alternative Procedures in Realizing of the High Frequency Power Standards with Microcalorimeter and Thermoelectric Power Sensors

The True-Twin microcalorimeter: a proof-of-concept experiment

Thermal Equivalence Error of Microwave Power Measurement

Improvements on INRIM Coaxial Microcalorimeter and Outcome of a Model Comparison

Power Sensor Characterization From 110 to 170 GHz Using a Waveguide Calorimeter

National Coaxial Primary Standard for Broadband Power

Measurement and Analysis for Performance of Micro-thermoelectric Generator

High-fidelity analysis of the electro-optical non-equivalence of planar electrical substitution radiometers

Multi-kilowatt cw laser power measurement comparison between national standards

Estimating Classification Uncertainty of Bayesian Decision Tree Technique on Financial Data

Cryogenic sensor enabling broad-band and traceable power measurements

Measurement of microwave photon correlations at millikelvin with a thermal detector

Empirical Evaluation Of Four Microwave Radiative Forward Models Based On Ground-Based Radiometer Data Near 20 And 30 Ghz

Thermal Emissivity Measurement with Two-temperature Method

SIM Thermocouple Type S Supplementary Comparison

Two Models for Bolometer and Microcalorimeter Detectors with Complex Thermal Architectures