Presentation 2021-03-05
A Proposal for Causal Inference with Subjective Evaluation
Daichi Ikeda, Hikaru Morita,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Machine learning techniques such as deep learning are often used for identification and personal matching in information security. In addition, an approach that increases the amount of data for training is generally used to improve the inference accuracy. However, in cases where a large amount of data is not available, the accuracy cannot be improved. In order to improve the inference accuracy, we propose a method to stimulate learning by adding subjective evaluation data without increasing the existing data. Specifically, we add subjective evaluation data as new random variables to the existing random variables consisting of features, etc. Here, for the interrelationship of the random variables, we took the approach of adding the random variables of subjective evaluation to the NBC (Naive Bayesian Classifier). In addition, by taking the same approach as the Probabilistic Graphical Model (hereinafter referred to as PGM), the formulation is formulated as an extension of NBC, so to speak. The validity of this method is evaluated as a problem of judging URL addresses for phishing and as a problem of judging conversational examples used in Telephone fraud.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Naive Bayesian Classifier / Probablistic Graphical Model / d-separataion
Paper # IT2020-146,ISEC2020-76,WBS2020-65
Date of Issue 2021-02-25 (IT, ISEC, WBS)

Conference Information
Committee WBS / IT / ISEC
Conference Date 2021/3/4(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Online
Topics (in Japanese) (See Japanese page)
Topics (in English) Joint Meeting of WBS, IT, and ISEC
Chair Masanori Hamamura(Kochi Univ. of Tech.) / Tadashi Wadayama(Nagoya Inst. of Tech.) / Shoichi Hirose(Univ. of Fukui)
Vice Chair Takashi Shono(INTEL) / Masahiro Fujii(Utsunomiya Univ.) / Tetsuya Kojima(Tokyo Kosen) / Tetsuya Izu(Fujitsu Labs.) / Noboru Kunihiro(Tsukuba Univ.)
Secretary Takashi Shono(Okayama Univ. of Science) / Masahiro Fujii(National Defence Academy) / Tetsuya Kojima(Yamaguchi Univ.) / Tetsuya Izu(Saga Univ.) / Noboru Kunihiro(Tsukuba Univ.)
Assistant Duong Quang Thang(NAIST) / Masafumi Moriyama(NICT) / Masayuki Kinoshita(Chiba Univ. of Tech.) / Takahiro Ohta(Senshu Univ.) / Kazuki Yoneyama(Ibaraki Univ.)

Paper Information
Registration To Technical Committee on Wideband System / Technical Committee on Information Theory / Technical Committee on Information Security
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Proposal for Causal Inference with Subjective Evaluation
Sub Title (in English)
Keyword(1) Naive Bayesian Classifier
Keyword(2) Probablistic Graphical Model
Keyword(3) d-separataion
1st Author's Name Daichi Ikeda
1st Author's Affiliation Graduate School of Kanagawa University(Graduate School of Kanagawa Univ.)
2nd Author's Name Hikaru Morita
2nd Author's Affiliation Graduate School of Kanagawa University(Graduate School of Kanagawa Univ.)
Date 2021-03-05
Paper # IT2020-146,ISEC2020-76,WBS2020-65
Volume (vol) vol.120
Number (no) IT-410,ISEC-411,WBS-412
Page pp.pp.208-212(IT), pp.208-212(ISEC), pp.208-212(WBS),
#Pages 5
Date of Issue 2021-02-25 (IT, ISEC, WBS)