Py-CIU: A Python Library for Explaining Machine Learning Predictions Using Contextual Importance and Utility

Sule Anjomshoae; Timotheus Kampik; Kary Främling

Py-CIU: A Python Library for Explaining Machine Learning Predictions Using Contextual Importance and Utility

Sule Anjomshoae, Timotheus Kampik, Kary Främling

Umeå University

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific

Abstract

In this paper, we present the Py-CIU library, a generic Python tool for applying the Contextual Importance and Utility (CIU) explainable machine learning method. CIU uses concepts from decision theory to explain a machine learning model’s prediction specific to a given data point by investigating the importance and usefulness of individual features (or feature combinations) to a prediction. The explanations aim to be intelligible to machine learning experts as well as non-technical users. The library can be applied to any black-box model that outputs a prediction value for all classes

Original language	English
Title of host publication	IJCAI-PRICAI 2020 Workshop on Explainable Artificial Intelligence (XAI)
Publisher	IJCAI
Publication status	Published - 2020
MoE publication type	B3 Non-refereed conference publication
Event	Workshop on Explainable Artificial Intelligence - Virtual, Online, Yokohama, Japan Duration: 8 Jan 2021 → 8 Jan 2021

Workshop

Workshop	Workshop on Explainable Artificial Intelligence
Country/Territory	Japan
City	Yokohama
Period	08/01/2021 → 08/01/2021

Access to Document

Cite this

@inproceedings{0932ced1949c4d938f18da2a60976d50,

title = "Py-CIU: A Python Library for Explaining Machine Learning Predictions Using Contextual Importance and Utility",

abstract = "In this paper, we present the Py-CIU library, a generic Python tool for applying the Contextual Importance and Utility (CIU) explainable machine learning method. CIU uses concepts from decision theory to explain a machine learning model{\textquoteright}s prediction specific to a given data point by investigating the importance and usefulness of individual features (or feature combinations) to a prediction. The explanations aim to be intelligible to machine learning experts as well as non-technical users. The library can be applied to any black-box model that outputs a prediction value for all classes",

author = "Sule Anjomshoae and Timotheus Kampik and Kary Fr{\"a}mling",

note = "Conference postponed from July 2020 to preliminary January 2021. ; Workshop on Explainable Artificial Intelligence ; Conference date: 08-01-2021 Through 08-01-2021",

year = "2020",

language = "English",

booktitle = "IJCAI-PRICAI 2020 Workshop on Explainable Artificial Intelligence (XAI)",

publisher = "IJCAI",

address = "United States",

}

Anjomshoae, S, Kampik, T & Främling, K 2020, Py-CIU: A Python Library for Explaining Machine Learning Predictions Using Contextual Importance and Utility. in IJCAI-PRICAI 2020 Workshop on Explainable Artificial Intelligence (XAI). IJCAI, Workshop on Explainable Artificial Intelligence, Yokohama, Japan, 08/01/2021. <http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-174782>

TY - GEN

T1 - Py-CIU: A Python Library for Explaining Machine Learning Predictions Using Contextual Importance and Utility

AU - Anjomshoae, Sule

AU - Kampik, Timotheus

AU - Främling, Kary

N1 - Conference postponed from July 2020 to preliminary January 2021.

PY - 2020

Y1 - 2020

N2 - In this paper, we present the Py-CIU library, a generic Python tool for applying the Contextual Importance and Utility (CIU) explainable machine learning method. CIU uses concepts from decision theory to explain a machine learning model’s prediction specific to a given data point by investigating the importance and usefulness of individual features (or feature combinations) to a prediction. The explanations aim to be intelligible to machine learning experts as well as non-technical users. The library can be applied to any black-box model that outputs a prediction value for all classes

AB - In this paper, we present the Py-CIU library, a generic Python tool for applying the Contextual Importance and Utility (CIU) explainable machine learning method. CIU uses concepts from decision theory to explain a machine learning model’s prediction specific to a given data point by investigating the importance and usefulness of individual features (or feature combinations) to a prediction. The explanations aim to be intelligible to machine learning experts as well as non-technical users. The library can be applied to any black-box model that outputs a prediction value for all classes

M3 - Conference article in proceedings

BT - IJCAI-PRICAI 2020 Workshop on Explainable Artificial Intelligence (XAI)

PB - IJCAI

T2 - Workshop on Explainable Artificial Intelligence

Y2 - 8 January 2021 through 8 January 2021

ER -

Py-CIU: A Python Library for Explaining Machine Learning Predictions Using Contextual Importance and Utility

Abstract

Workshop

Access to Document

Fingerprint

Cite this