Functional polymers are essential materials supporting modern society and are being actively researched in an experiment-centric manner. Use of artificial intelligence (AI) or machine learning (ML) are expected to further accelerate research efficiency, but low transparency and interpretability of AI deter researchers from trusting it. This study built an explainable AI (XAI) to predict property of anion exchange membrane, a kind of functional polymer. This study is conducted in four steps: 1. Construction of an in-house database (DB); 2. Digitization of polymer structure using existing descriptors; 3. Construction of an ML model; 4. Calculate and analyze the Shapley (SHAP) values for each explanatory variable for evaluating explainability and transparency. Open DB is not available for the target material in this study, hence an in-house DB consisting structural property data of around 300 polymers was built. From 2. and 3., we built an ML model with test data prediction accuracy of 0.7983. Carrying out 4., we found that AMID_N, a descriptor-origin explanatory variable highly correlating polymer substructure and its property, is important. The findings of such important features strongly support chemical interpretation, thereby successfully obtaining a XAI that can be used in experiment cycle.
"},{"@language":"ja","@value":"現代社会を支える基幹材料である機能性高分子は、実験を中心に盛んに研究されているが、人工知能（AI）や機械学習（ML）の導入による更なる効率化が期待されている。しかしながら、透明性・解釈性の低いMLモデルでは実験研究者からの信頼を得にくい。本研究では、機能性高分子材料の1つであるアニオン伝導膜の物性予測のための説明可能AI(XAI)を構築し、その透明性・解釈性を評価した。当研究の流れは①独自データベース(DB)の構築、②既存記述子を用いた高分子構造の数値化、③MLモデルの構築、④各説明変数のShapley(SHAP)値を算出し解析した。①本研究で対象としたアニオン伝導膜の構造・物性については公開DBが存在しないため、300弱の構造・物性データを論文から収集してDB化した。次に、②と③を実施した結果、テストデータに対してR2=0.7983の予測精度を持つモデルを得た。④では高分子構造記述子由来のAMID_Nが重要であることが示された。AMID_Nは高分子構造内の物性と密に相関する記述子であり、化学的な解釈・理解を後押しし、実験へのフィードバックが可能なXAIの構築に成功した。
"}],"abstractLicenseFlag":"disallow"}],"creator":[{"@id":"https://cir.nii.ac.jp/crid/1410863395972104321","@type":"Researcher","foaf:name":[{"@language":"en","@value":"PHUA Yin Kan"},{"@language":"ja","@value":"PHUA Yin Kan"}],"jpcoar:affiliationName":[{"@language":"ja","@value":"九州大学大学院工学府"},{"@language":"en","@value":"Grad. Sch. of Eng., Kyushu University"}]},{"@id":"https://cir.nii.ac.jp/crid/1410863395972104320","@type":"Researcher","foaf:name":[{"@language":"ja","@value":"藤ヶ谷 剛彦"},{"@language":"en","@value":"FUJIGAYA Tsuyohiko"}],"jpcoar:affiliationName":[{"@language":"ja","@value":"九州大学大学院工学府"},{"@language":"en","@value":"Grad. Sch. of Eng., Kyushu University"},{"@language":"ja","@value":"九州大学分子システム科学センター"},{"@language":"en","@value":"CMS, Kyushu University"},{"@language":"en","@value":"I2CNER, Kyushu University"},{"@language":"ja","@value":"九州大学カーボンニュートラル・エネルギー国際研究所"}]},{"@id":"https://cir.nii.ac.jp/crid/1410863395972104322","@type":"Researcher","foaf:name":[{"@language":"en","@value":"KATO Koichiro"},{"@language":"ja","@value":"加藤 幸一郎"}],"jpcoar:affiliationName":[{"@language":"ja","@value":"九州大学大学院工学府"},{"@language":"en","@value":"Grad. Sch. of Eng., Kyushu University"},{"@language":"en","@value":"CMS, Kyushu University"},{"@language":"ja","@value":"九州大学分子システム科学センター"},{"@language":"ja","@value":"九州大学情報基盤研究開発センター"},{"@language":"en","@value":"RIIT, Kyushu University"}]}],"publication":{"publicationIdentifier":[{"@type":"EISSN","@value":"27587347"}],"prism:publicationName":[{"@language":"en","@value":"Proceedings of the Annual Conference of JSAI"},{"@language":"ja","@value":"人工知能学会全国大会論文集"},{"@language":"en","@value":"Proc. of JSAI"},{"@language":"ja","@value":"JSAI大会論文集"}],"dc:publisher":[{"@language":"en","@value":"The Japanese Society for Artificial Intelligence"},{"@language":"ja","@value":"一般社団法人 人工知能学会"}],"prism:publicationDate":"2024","prism:volume":"JSAI2024","prism:number":"0","prism:startingPage":"1F4GS1002","prism:endingPage":"1F4GS1002"},"jpcoar:conferenceName":"2024年度人工知能学会全国大会（第38回）","jpcoar:conferencePlace":"アクトシティ浜松＋オンライン","availableAt":"2024","foaf:topic":[{"@id":"https://cir.nii.ac.jp/all?q=%E9%AB%98%E5%88%86%E5%AD%90","dc:title":"高分子"},{"@id":"https://cir.nii.ac.jp/all?q=%E3%83%9E%E3%83%86%E3%83%AA%E3%82%A2%E3%83%AB%E3%82%BA%E3%82%A4%E3%83%B3%E3%83%95%E3%82%A9%E3%83%BC%E3%83%9E%E3%83%86%E3%82%A3%E3%82%AF%E3%82%B9","dc:title":"マテリアルズインフォーマティクス"},{"@id":"https://cir.nii.ac.jp/all?q=%E7%87%83%E6%96%99%E9%9B%BB%E6%B1%A0","dc:title":"燃料電池"},{"@id":"https://cir.nii.ac.jp/all?q=Polymer","dc:title":"Polymer"},{"@id":"https://cir.nii.ac.jp/all?q=Materials%20informatic","dc:title":"Materials informatic"},{"@id":"https://cir.nii.ac.jp/all?q=Fuel%20cell","dc:title":"Fuel cell"}],"dataSourceIdentifier":[{"@type":"JALC","@value":"oai:japanlinkcenter.org:2013117686"}]}