BTSC: Binary tree structure convolution layers for building interpretable decision-making deep CNN

IF 8.4 2区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE CAAI Transactions on Intelligence Technology Pub Date : 2024-03-31 DOI:10.1049/cit2.12328
Yuqi Wang, Dawei Dai, Da Liu, Shuyin Xia, Guoyin Wang
{"title":"BTSC: Binary tree structure convolution layers for building interpretable decision-making deep CNN","authors":"Yuqi Wang,&nbsp;Dawei Dai,&nbsp;Da Liu,&nbsp;Shuyin Xia,&nbsp;Guoyin Wang","doi":"10.1049/cit2.12328","DOIUrl":null,"url":null,"abstract":"<p>Although deep convolution neural network (DCNN) has achieved great success in computer vision field, such models are considered to lack interpretability in decision-making. One of fundamental issues is that its decision mechanism is considered to be a “black-box” operation. The authors design the binary tree structure convolution (BTSC) module and control the activation level of particular neurons to build the interpretable DCNN model. First, the authors design a BTSC module, in which each parent node generates two independent child layers, and then integrate them into a normal DCNN model. The main advantages of the BTSC are as follows: 1) child nodes of the different parent nodes do not interfere with each other; 2) parent and child nodes can inherit knowledge. Second, considering the activation level of neurons, the authors design an information coding objective to guide neural nodes to learn the particular information coding that is expected. Through the experiments, the authors can verify that: 1) the decision-making made by both the ResNet and DenseNet models can be explained well based on the \"decision information flow path\" (known as <b>the decision-path</b>) formed in the BTSC module; 2) <b>the decision-path</b> can reasonably interpret the decision reversal mechanism (Robustness mechanism) of the DCNN model; 3) the credibility of decision-making can be measured by the matching degree between the actual and expected <b>decision-path</b>.</p>","PeriodicalId":46211,"journal":{"name":"CAAI Transactions on Intelligence Technology","volume":"9 5","pages":"1331-1345"},"PeriodicalIF":8.4000,"publicationDate":"2024-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/cit2.12328","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"CAAI Transactions on Intelligence Technology","FirstCategoryId":"94","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1049/cit2.12328","RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0

Abstract

Although deep convolution neural network (DCNN) has achieved great success in computer vision field, such models are considered to lack interpretability in decision-making. One of fundamental issues is that its decision mechanism is considered to be a “black-box” operation. The authors design the binary tree structure convolution (BTSC) module and control the activation level of particular neurons to build the interpretable DCNN model. First, the authors design a BTSC module, in which each parent node generates two independent child layers, and then integrate them into a normal DCNN model. The main advantages of the BTSC are as follows: 1) child nodes of the different parent nodes do not interfere with each other; 2) parent and child nodes can inherit knowledge. Second, considering the activation level of neurons, the authors design an information coding objective to guide neural nodes to learn the particular information coding that is expected. Through the experiments, the authors can verify that: 1) the decision-making made by both the ResNet and DenseNet models can be explained well based on the "decision information flow path" (known as the decision-path) formed in the BTSC module; 2) the decision-path can reasonably interpret the decision reversal mechanism (Robustness mechanism) of the DCNN model; 3) the credibility of decision-making can be measured by the matching degree between the actual and expected decision-path.

Abstract Image

查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
BTSC:用于构建可解释决策深度 CNN 的二叉树结构卷积层
虽然深度卷积神经网络(DCNN)在计算机视觉领域取得了巨大成功,但这类模型被认为在决策方面缺乏可解释性。其中一个根本问题是其决策机制被认为是 "黑箱 "操作。作者设计了二叉树结构卷积(BTSC)模块,并控制特定神经元的激活水平,以建立可解释的 DCNN 模型。首先,作者设计了一个 BTSC 模块,其中每个父节点生成两个独立的子层,然后将它们集成到一个普通的 DCNN 模型中。BTSC 的主要优点如下:1)不同父节点的子节点互不干扰;2)父节点和子节点可以继承知识。其次,考虑到神经元的激活水平,作者设计了一个信息编码目标,引导神经节点学习预期的特定信息编码。通过实验,作者可以验证1)根据 BTSC 模块中形成的 "决策信息流路径"(即决策路径),可以很好地解释 ResNet 和 DenseNet 模型的决策;2)决策路径可以合理地解释 DCNN 模型的决策逆转机制(鲁棒性机制);3)决策的可信度可以通过实际决策路径与预期决策路径的匹配程度来衡量。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
CAAI Transactions on Intelligence Technology
CAAI Transactions on Intelligence Technology COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE-
CiteScore
11.00
自引率
3.90%
发文量
134
审稿时长
35 weeks
期刊介绍: CAAI Transactions on Intelligence Technology is a leading venue for original research on the theoretical and experimental aspects of artificial intelligence technology. We are a fully open access journal co-published by the Institution of Engineering and Technology (IET) and the Chinese Association for Artificial Intelligence (CAAI) providing research which is openly accessible to read and share worldwide.
期刊最新文献
Guest Editorial: Knowledge-based deep learning system in bio-medicine Guest Editorial: Special issue on trustworthy machine learning for behavioural and social computing A fault-tolerant and scalable boosting method over vertically partitioned data Multi-objective interval type-2 fuzzy linear programming problem with vagueness in coefficient Prediction and optimisation of gasoline quality in petroleum refining: The use of machine learning model as a surrogate in optimisation framework
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1