%0 Conference Proceedings
%T Reliable Probability Estimates Based on Support Vector Machines for Large Multiclass Datasets
%+ Department of Computer Science (Royal Holloway University of London)
%+ Computer Science and Engineering Department
%+ Computer Learning Research Centre, Royal Holloway
%A Lambrou, Antonis
%A Papadopoulos, Harris
%A Nouretdinov, Ilia
%A Gammerman, Alexander
%Z Part 4: First Conformal Prediction and Its Applications Workshop (COPA 2012)
%< avec comité de lecture
%( IFIP Advances in Information and Communication Technology
%B 8th International Conference on Artificial Intelligence Applications and Innovations (AIAI)
%C Halkidiki, Greece
%Y Lazaros Iliadis
%Y Ilias Maglogiannis
%Y Harris Papadopoulos
%Y Kostas Karatzas
%Y Spyros Sioutas
%I Springer
%3 Artificial Intelligence Applications and Innovations
%V AICT-382
%N Part II
%P 182-191
%8 2012-09-27
%D 2012
%R 10.1007/978-3-642-33412-2_19
%K Support Vector Machine
%K well calibrated probabilities
%K multiclass
%K Inductive Venn Predictor
%K Machine Learning
%Z Computer Science [cs]Conference papers
%X Venn Predictors (VPs) are machine learning algorithms that can provide well calibrated multiprobability outputs for their predictions. The only drawback of Venn Predictors is their computational inefficiency, especially in the case of large datasets. In this work, we propose an Inductive Venn Predictor (IVP) which overcomes the computational inefficiency problem of the original Venn Prediction framework. Each VP is defined by a taxonomy which separates the data into categories. We develop an IVP with a taxonomy derived from a multiclass Support Vector Machine (SVM), and we compare our method with other probabilistic methods for SVMs, namely Platt’s method, SVM Binning, and SVM with Isotonic Regression. We show that these methods do not always provide well calibrated outputs, while our IVP will always guarantee this property under the i.i.d. assumption.
%G English
%Z TC 12
%Z WG 12.5
%2 https://hal.science/hal-01523046/document
%2 https://hal.science/hal-01523046/file/978-3-642-33412-2_19_Chapter.pdf
%L hal-01523046
%U https://hal.science/hal-01523046
%~ IFIP
%~ IFIP-AICT
%~ IFIP-TC
%~ IFIP-WG
%~ IFIP-TC12
%~ IFIP-AIAI
%~ IFIP-WG12-5
%~ IFIP-AICT-382