Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00217
Andreas Mäkilä, A. Friebe, Leif Enblom, P. Strandberg, T. Seceleanu
Many hardware solutions for Power over Ethernet (PoE) Power Sourcing Equipment (PSE) exist, with slightly varying feature sets. A software solution is needed for interaction with the PSEs, and for managing a power budget across several PSEs. A generic interface is desirable, as well as generic software components that can be used in support of several PSE solutions. In this paper we present a union of features and real-time requirements for three hardware solutions, and the development of a generic software architecture.
{"title":"A Generic Software Architecture for PoE Power Sourcing Equipment","authors":"Andreas Mäkilä, A. Friebe, Leif Enblom, P. Strandberg, T. Seceleanu","doi":"10.1109/COMPSAC54236.2022.00217","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00217","url":null,"abstract":"Many hardware solutions for Power over Ethernet (PoE) Power Sourcing Equipment (PSE) exist, with slightly varying feature sets. A software solution is needed for interaction with the PSEs, and for managing a power budget across several PSEs. A generic interface is desirable, as well as generic software components that can be used in support of several PSE solutions. In this paper we present a union of features and real-time requirements for three hardware solutions, and the development of a generic software architecture.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"131 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127035034","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00240
Lili Xiao, Huibiao Zhu, Mengda He, S. Qin
The C++11 standard introduced a language level weak memory model (i.e., the C++11 memory model) to improve the performance of the execution of C/C++ programs. Algebra is well-suited for direct use by engineers in symbolic calculation of parameters. It is a challenge to investigate the algebraic semantics for the C++11 memory model. Inspired by the promising semantics, in this paper, we explore the algebraic laws for the C++11 memory model, including a set of sequential and parallel expansion laws. We introduce the concept of guarded choice, and every program under the C++11 memory model can be converted into the head normal form of guarded choice. In addition, the proposed algebraic laws are implemented in the rewriting engine Maude.
c++ 11标准引入了一种语言级弱内存模型(即c++ 11内存模型)，以提高C/ c++程序的执行性能。代数非常适合工程师在参数的符号计算中直接使用。研究c++ 11内存模型的代数语义是一个挑战。受这一前景语义的启发，本文探讨了c++ 11内存模型的代数定律，包括一组顺序和并行展开定律。引入了保护选择的概念，c++ 11内存模型下的每个程序都可以转换为保护选择的头部范式。此外，所提出的代数定律在重写引擎Maude中实现。
{"title":"Algebraic Semantics for C++11 Memory Model","authors":"Lili Xiao, Huibiao Zhu, Mengda He, S. Qin","doi":"10.1109/COMPSAC54236.2022.00240","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00240","url":null,"abstract":"The C++11 standard introduced a language level weak memory model (i.e., the C++11 memory model) to improve the performance of the execution of C/C++ programs. Algebra is well-suited for direct use by engineers in symbolic calculation of parameters. It is a challenge to investigate the algebraic semantics for the C++11 memory model. Inspired by the promising semantics, in this paper, we explore the algebraic laws for the C++11 memory model, including a set of sequential and parallel expansion laws. We introduce the concept of guarded choice, and every program under the C++11 memory model can be converted into the head normal form of guarded choice. In addition, the proposed algebraic laws are implemented in the rewriting engine Maude.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127165144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00107
Natsumi Matsui, Ayumi Ohnishi, T. Terada, M. Tsukamoto
Curling irons can create curled hair by heating it. However, because the operation of the curling iron involves six degrees of freedom (6DoF) movements, such as translation and rotation, it is difficult to understand and reproduce by simply watching instructional videos. In this study, we proposed Color-Path, a smart mirror-shaped system that allows users to easily understand how to move a curling iron. The proposed system acquires the moving paths of the curling iron using a camera on a smart mirror and an accelerometer on the curling iron. The system displays the curling iron paths of the target hairstyle on a smart mirror in Augmented Reality (AR). From the evaluation experiments, we confirmed that the system contributed to the reproduction of the moving paths of the curling iron through quantitative evaluation. However, a subjective evaluation showed that our system could not reproduce the appearance of the target hairstyle. The results indicate that the time to heat the hair should be considered.
{"title":"Color-Path: Hair Arrangement Reproduction Support System by Displaying Target Motion in AR","authors":"Natsumi Matsui, Ayumi Ohnishi, T. Terada, M. Tsukamoto","doi":"10.1109/COMPSAC54236.2022.00107","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00107","url":null,"abstract":"Curling irons can create curled hair by heating it. However, because the operation of the curling iron involves six degrees of freedom (6DoF) movements, such as translation and rotation, it is difficult to understand and reproduce by simply watching instructional videos. In this study, we proposed Color-Path, a smart mirror-shaped system that allows users to easily understand how to move a curling iron. The proposed system acquires the moving paths of the curling iron using a camera on a smart mirror and an accelerometer on the curling iron. The system displays the curling iron paths of the target hairstyle on a smart mirror in Augmented Reality (AR). From the evaluation experiments, we confirmed that the system contributed to the reproduction of the moving paths of the curling iron through quantitative evaluation. However, a subjective evaluation showed that our system could not reproduce the appearance of the target hairstyle. The results indicate that the time to heat the hair should be considered.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"538 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124254810","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00190
Yoshihiro Tsuboki, Tomoya Kawakami, Satoru Matsumoto, T. Yoshihisa, Y. Teranishi
Recent technological advances in Virtual Reality (VR) and Augmented Reality (AR) enable users to experience a high-quality virtual world. In VR applications, the user's physical movement is generally restricted because the situation around the real world cannot be seen. AR allows users to experience virtual worlds without restrictions on physical movement, but the extent to which they are replaced as virtual worlds is limited. In this research, assuming the use of smartphones and tablet devices, a partial virtual world system is implemented by removing only the background part from the real-time real-world image taken by the camera and replacing it with a virtual background.
{"title":"A Real-Time Background Replacement Method Based on Estimated Depth for AR Applications","authors":"Yoshihiro Tsuboki, Tomoya Kawakami, Satoru Matsumoto, T. Yoshihisa, Y. Teranishi","doi":"10.1109/COMPSAC54236.2022.00190","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00190","url":null,"abstract":"Recent technological advances in Virtual Reality (VR) and Augmented Reality (AR) enable users to experience a high-quality virtual world. In VR applications, the user's physical movement is generally restricted because the situation around the real world cannot be seen. AR allows users to experience virtual worlds without restrictions on physical movement, but the extent to which they are replaced as virtual worlds is limited. In this research, assuming the use of smartphones and tablet devices, a partial virtual world system is implemented by removing only the background part from the real-time real-world image taken by the camera and replacing it with a virtual background.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124353996","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00171
A. Rahman, Farzana Ahamed Bhuiyan, M. M. Hassan, H. Shahriar, Fan Wu
Machine learning (ML) operations or MLOps advo-cates for integration of DevOps- related practices into the ML development and deployment process. Adoption of MLOps can be hampered due to a lack of knowledge related to how development tasks can be automated. A characterization of bot usage in ML projects can help practitioners on the types of tasks that can be automated with bots, and apply that knowledge into their ML development and deployment process. To that end, we conduct a preliminary empirical study with 135 issues reported mined from 3 libraries related to deep learning: Keras, PyTorch, and Tensorflow. From our empirical study we observe 9 categories of tasks that are automated with bots. We conclude our work-in-progress paper by providing a list of lessons that we learned from our empirical study.
{"title":"Towards Automation for MLOps: An Exploratory Study of Bot Usage in Deep Learning Libraries","authors":"A. Rahman, Farzana Ahamed Bhuiyan, M. M. Hassan, H. Shahriar, Fan Wu","doi":"10.1109/COMPSAC54236.2022.00171","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00171","url":null,"abstract":"Machine learning (ML) operations or MLOps advo-cates for integration of DevOps- related practices into the ML development and deployment process. Adoption of MLOps can be hampered due to a lack of knowledge related to how development tasks can be automated. A characterization of bot usage in ML projects can help practitioners on the types of tasks that can be automated with bots, and apply that knowledge into their ML development and deployment process. To that end, we conduct a preliminary empirical study with 135 issues reported mined from 3 libraries related to deep learning: Keras, PyTorch, and Tensorflow. From our empirical study we observe 9 categories of tasks that are automated with bots. We conclude our work-in-progress paper by providing a list of lessons that we learned from our empirical study.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129124968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00185
Adrian F. Spataru, Gabriel Iuhasz, S. Panica
A Distributed Application Topology is a valuable commodity built on the strength of a long and iterative design process. A topology is generally refined over time, other topologies can use it as a component, and the community may share it. To reproduce a deployment, several properties must be recorded such as data origin, processing steps, configuration settings, and hardware requirements. This paper presents an extension to the TOSCA specification that allows for the definition of accelerator-aware services that can span from Cloud to Edge. Additionally, we introduce the concept of Abstract Applications that contain at least one abstract service definition. The process of Service Optimization replaces the abstract sertvices, creating an explicit topology deployable under hybrid deployment models (Virtual Machines, Containers, HPC) residing on the Cloud Continuum spectrum.
{"title":"TUFA: A TOSCA extension for the specification of accelerator-aware applications in the Cloud Continuum","authors":"Adrian F. Spataru, Gabriel Iuhasz, S. Panica","doi":"10.1109/COMPSAC54236.2022.00185","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00185","url":null,"abstract":"A Distributed Application Topology is a valuable commodity built on the strength of a long and iterative design process. A topology is generally refined over time, other topologies can use it as a component, and the community may share it. To reproduce a deployment, several properties must be recorded such as data origin, processing steps, configuration settings, and hardware requirements. This paper presents an extension to the TOSCA specification that allows for the definition of accelerator-aware services that can span from Cloud to Edge. Additionally, we introduce the concept of Abstract Applications that contain at least one abstract service definition. The process of Service Optimization replaces the abstract sertvices, creating an explicit topology deployable under hybrid deployment models (Virtual Machines, Containers, HPC) residing on the Cloud Continuum spectrum.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"100 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133690184","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00241
Chawanangwa Lupafya, D. Balasubramaniam
There are many aspects involved in the development and operation of a software system, including system artefacts, activities, and infrastructure. Most of these aspects are vulnerable to uncertainty, which can result in risks to system quality and performance. Thus it is important to identify, represent and manage uncertainty in software systems. We hypothesise that using an underlying conceptual framework for characterising uncertainty can facilitate these activities. This paper demonstrates the use of an extensible framework, which defines a foundation for the systematic and explicit consideration of uncertainty in software systems. A software architecture case study is used to illustrate and evaluate the framework. A discussion of potential uses for the framework and future research is also provided.
{"title":"A Framework for Considering Uncertainty in Software Systems","authors":"Chawanangwa Lupafya, D. Balasubramaniam","doi":"10.1109/COMPSAC54236.2022.00241","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00241","url":null,"abstract":"There are many aspects involved in the development and operation of a software system, including system artefacts, activities, and infrastructure. Most of these aspects are vulnerable to uncertainty, which can result in risks to system quality and performance. Thus it is important to identify, represent and manage uncertainty in software systems. We hypothesise that using an underlying conceptual framework for characterising uncertainty can facilitate these activities. This paper demonstrates the use of an extensible framework, which defines a foundation for the systematic and explicit consideration of uncertainty in software systems. A software architecture case study is used to illustrate and evaluate the framework. A discussion of potential uses for the framework and future research is also provided.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130368362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00035
Xin Xu, Nan Wang, Huijie Jin, Yang Liu, Kun Li
The rapid development of neural networks has con-tributed to the increasing maturity of recommendation systems. However, deep neural networks have poor interpretability for models and do not show strong advantages for sparse data and noisy data. Recently, Hawkes process has become more and more focused for its good interpretability with probabilistic models. Based on this, we proposes A Neural Efficient Recommendation Model Based on Neighborhood Key Information Aggregation of Modified Hawkes(NKMH). The model utilizes a neural network and designs three modules to jointly fit the modified Hawkes process. It not only inherits the high interpretability of Hawkes, but also effectively solves the problem of poor prediction ability of the Hawkes process. Besides, we present a novel key information search strategy(KISS), which can effectively remove the noise in a session and alleviate the sparsity of the data to some extent. Extensive experiments on two datasets show that the NKMH model outperforms many current popular models.
{"title":"NKMH: A Neural Efficient Recommendation Based on Neighborhood Key Information Aggregation of Modified Hawkes","authors":"Xin Xu, Nan Wang, Huijie Jin, Yang Liu, Kun Li","doi":"10.1109/COMPSAC54236.2022.00035","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00035","url":null,"abstract":"The rapid development of neural networks has con-tributed to the increasing maturity of recommendation systems. However, deep neural networks have poor interpretability for models and do not show strong advantages for sparse data and noisy data. Recently, Hawkes process has become more and more focused for its good interpretability with probabilistic models. Based on this, we proposes A Neural Efficient Recommendation Model Based on Neighborhood Key Information Aggregation of Modified Hawkes(NKMH). The model utilizes a neural network and designs three modules to jointly fit the modified Hawkes process. It not only inherits the high interpretability of Hawkes, but also effectively solves the problem of poor prediction ability of the Hawkes process. Besides, we present a novel key information search strategy(KISS), which can effectively remove the noise in a session and alleviate the sparsity of the data to some extent. Extensive experiments on two datasets show that the NKMH model outperforms many current popular models.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"45 17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130435459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-01DOI: 10.1109/COMPSAC54236.2022.00154
Wei Liu, Jiarui Zhang, Yijun Zhao
Deep learning has attracted a great amount of interest in recent years and has become a rapidly emerging field in artificial intelligence. In medical image analysis, deep learning methods have produced promising results comparable to and, in some cases, superior to human experts. Nevertheless, researchers have also noted the limitations and challenges of the deep learning approaches, especially in model selection and interpretability. This paper compares the efficacy of deep learning and traditional machine learning techniques in detecting cognitive impairment (CI) associated with Alzheimer's disease (AD) using brain MRI scans. We base our study on 894 brain MRI scans provided by the open access OASIS platform. In particular, we explore two deep learning approaches: 1) a 3D convolutional neural network (3D-CNN) and 2) a hybrid model with a CNN plus LSTM (CNN-LSTM) architecture. We further examine the performance of five traditional machine learning algorithms based on features extracted from the MRI images using the FreeSurfer software. Our experimental results demonstrate that the deep learning models achieve higher Precision and Recall, while the traditional machine learning methods deliver more stability and better performance in Specificity and overall accuracy. Our findings could serve as a case study to highlight the challenges in adopting deep learning-based approaches.
{"title":"A Comparison of Deep Learning and Traditional Machine Learning Approaches in Detecting Cognitive Impairment Using MRI Scans","authors":"Wei Liu, Jiarui Zhang, Yijun Zhao","doi":"10.1109/COMPSAC54236.2022.00154","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00154","url":null,"abstract":"Deep learning has attracted a great amount of interest in recent years and has become a rapidly emerging field in artificial intelligence. In medical image analysis, deep learning methods have produced promising results comparable to and, in some cases, superior to human experts. Nevertheless, researchers have also noted the limitations and challenges of the deep learning approaches, especially in model selection and interpretability. This paper compares the efficacy of deep learning and traditional machine learning techniques in detecting cognitive impairment (CI) associated with Alzheimer's disease (AD) using brain MRI scans. We base our study on 894 brain MRI scans provided by the open access OASIS platform. In particular, we explore two deep learning approaches: 1) a 3D convolutional neural network (3D-CNN) and 2) a hybrid model with a CNN plus LSTM (CNN-LSTM) architecture. We further examine the performance of five traditional machine learning algorithms based on features extracted from the MRI images using the FreeSurfer software. Our experimental results demonstrate that the deep learning models achieve higher Precision and Recall, while the traditional machine learning methods deliver more stability and better performance in Specificity and overall accuracy. Our findings could serve as a case study to highlight the challenges in adopting deep learning-based approaches.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117215111","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Credit risk assessment, which aims at identifying high-risk users, plays a critical role in financial institutions. A common method is to use the greedy strategy to generate an interpretable rule set to classify all the users into high-risk or non-risk users. During each iteration, the greedy strategy utilizes a pre-defined indicator function to evaluate which rule is the best and then adds it to the rule set. However, in reality, the indicator function is designed manually and requires much domain knowledge and expert experience. Worse still, we need to design a suitable indicator for every situation, which is tedious and time-consuming work. This motivates us to propose a self-adaptive indicator that can be adapted to different situations without too much human intervention. In this paper, we see the indicator as a weighted sum of several sub-indicators. By tuning the weights, the indicator can be adapted to different situations automatically. That is, we transform this indicator selection problem into a weights tuning problem. To find the best weight of self-adaptive indicators, machine learning methods and black-box optimization are utilized. The experimental results demonstrated that our self-adaptive indicator can select a better rule set to identify more high-risk users compared to the human-defined indicator.
{"title":"A Self-adaptive Indicator Selection Approach for Solving Credit Risk Assessment","authors":"Yongfeng Gu, Yue Ning, Hao Ding, Kecai Gu, Daohong Jian, Zhou Xu, Hua Wu, Jun Zhou","doi":"10.1109/COMPSAC54236.2022.00249","DOIUrl":"https://doi.org/10.1109/COMPSAC54236.2022.00249","url":null,"abstract":"Credit risk assessment, which aims at identifying high-risk users, plays a critical role in financial institutions. A common method is to use the greedy strategy to generate an interpretable rule set to classify all the users into high-risk or non-risk users. During each iteration, the greedy strategy utilizes a pre-defined indicator function to evaluate which rule is the best and then adds it to the rule set. However, in reality, the indicator function is designed manually and requires much domain knowledge and expert experience. Worse still, we need to design a suitable indicator for every situation, which is tedious and time-consuming work. This motivates us to propose a self-adaptive indicator that can be adapted to different situations without too much human intervention. In this paper, we see the indicator as a weighted sum of several sub-indicators. By tuning the weights, the indicator can be adapted to different situations automatically. That is, we transform this indicator selection problem into a weights tuning problem. To find the best weight of self-adaptive indicators, machine learning methods and black-box optimization are utilized. The experimental results demonstrated that our self-adaptive indicator can select a better rule set to identify more high-risk users compared to the human-defined indicator.","PeriodicalId":330838,"journal":{"name":"2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC)","volume":"102 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131438804","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：