Online ISSN: 1884-4111 Print ISSN: 0033-8303
Radioisotopes 66(1): 1-10 (2017)


ランダムフォレスト法によるγ線スペクトルを用いた放射性廃棄物ドラム缶の分類Classification of Radioactive Waste Drums Using Random Forests for Their γ-ray Spectra

1国立研究開発法人日本原子力研究開発機構バックエンド研究開発部門人形峠環境技術センターNingyo-toge Environmental Engineering Center, 
Sector of Decommissioning and Radioactive Waste Management, Japan Atomic Energy Agency ◇ 708–0698 岡山県苫田郡鏡野町上齋原1550 ◇ 1550 Kamisaibara, Kagamino-cho, Tomata-gun, Okayama Pref. 708–0698, Japan

2岡山大学大学院自然科学研究科Graduate School of Natural Science and Technology, Okayama University ◇ 700–8530 岡山県北区津島中3–1–1 ◇ 3–1–1 Tsushima-naka, Kita-ku, Okayama Pref. 700–8530, Japan

受付日:2016年7月4日Received: July 4, 2016
受理日:2016年9月9日Accepted: September 9, 2016
発行日:2017年1月15日Published: January 15, 2017


The feasibility of Random Forests, one of machine learning methods was examined for the classification of radioactive waste drums. It was carried out using 954 γ-ray spectra of drums which were already classified to natural or reprocessed uranium. After 300 spectra were selected at random to reassemble training datasets, the percentages of correct classification by Random Forests were evaluated with another 654 spectra. When the counts of spectra were reprocessed as the difference of their logarithm, Random Forests accurately classified 654 drums.

Key words: radioactive waste; machine learning; Random Forests; γ-ray measurement

This page was created on 2016-11-24T09:21:32.204+09:00
This page was last modified on 2017-01-06T15:00:58.495+09:00