Reverse Conformal Approach for On-line Experimental Design. / Nouretdinov, Ilia.

6th Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2017). Vol. 60 2017. p. 185-192.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Published

Standard

Reverse Conformal Approach for On-line Experimental Design. / Nouretdinov, Ilia.

6th Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2017). Vol. 60 2017. p. 185-192.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Harvard

Nouretdinov, I 2017, Reverse Conformal Approach for On-line Experimental Design. in 6th Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2017). vol. 60, pp. 185-192. <http://proceedings.mlr.press/v60/nouretdinov17a.html>

APA

Nouretdinov, I. (2017). Reverse Conformal Approach for On-line Experimental Design. In 6th Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2017) (Vol. 60, pp. 185-192) http://proceedings.mlr.press/v60/nouretdinov17a.html

Vancouver

Nouretdinov I. Reverse Conformal Approach for On-line Experimental Design. In 6th Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2017). Vol. 60. 2017. p. 185-192

Author

Nouretdinov, Ilia. / Reverse Conformal Approach for On-line Experimental Design. 6th Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2017). Vol. 60 2017. pp. 185-192

BibTeX

@inproceedings{72cececee8f741329e1aa779601d92d9,
title = "Reverse Conformal Approach for On-line Experimental Design",
abstract = "Conformal prediction is a recently developed framework of confident machine learning with guaranteed validity properties for prediction sets. In this work we study its usage in reversed version of the traditional machine learning problem: prediction of objects which can have a given label, instead of usual prediction of labels by objects. It is meant that the label reflect some desired property of the object. For this kind of task, the conformal prediction framework can provide a prediction set that is a set of objects that are likely to have the label. Based on this, we create an on-line protocol of experimental design. It includes a choice criterion based on conformal output, and elements of transfer learning in order to keep the validity properties in on-line regime. ",
keywords = "Confident classification, conformal prediction, experimental design, transfer learning",
author = "Ilia Nouretdinov",
year = "2017",
month = jun,
language = "English",
volume = "60",
pages = "185--192",
booktitle = "6th Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2017)",

}

RIS

TY - GEN

T1 - Reverse Conformal Approach for On-line Experimental Design

AU - Nouretdinov, Ilia

PY - 2017/6

Y1 - 2017/6

N2 - Conformal prediction is a recently developed framework of confident machine learning with guaranteed validity properties for prediction sets. In this work we study its usage in reversed version of the traditional machine learning problem: prediction of objects which can have a given label, instead of usual prediction of labels by objects. It is meant that the label reflect some desired property of the object. For this kind of task, the conformal prediction framework can provide a prediction set that is a set of objects that are likely to have the label. Based on this, we create an on-line protocol of experimental design. It includes a choice criterion based on conformal output, and elements of transfer learning in order to keep the validity properties in on-line regime.

AB - Conformal prediction is a recently developed framework of confident machine learning with guaranteed validity properties for prediction sets. In this work we study its usage in reversed version of the traditional machine learning problem: prediction of objects which can have a given label, instead of usual prediction of labels by objects. It is meant that the label reflect some desired property of the object. For this kind of task, the conformal prediction framework can provide a prediction set that is a set of objects that are likely to have the label. Based on this, we create an on-line protocol of experimental design. It includes a choice criterion based on conformal output, and elements of transfer learning in order to keep the validity properties in on-line regime.

KW - Confident classification, conformal prediction, experimental design, transfer learning

M3 - Conference contribution

VL - 60

SP - 185

EP - 192

BT - 6th Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2017)

ER -