Pub Date : 2021-10-15DOI: 10.1109/CCCI52664.2021.9583199
Le Weng, Hengyu Liu, Lianfeng Huang, Yingmin Zhang, Chao Feng
With the vigorous development of the mobile Internet, the Android system, which accounts for 76% of the drama mobile operating system, has also been widely promoted and popularized. However, due to its own open source characteristics, various parts of the Android system are facing serious threats from hacker attacks, and the main threat comes from malicious applications. In order to cope with the challenge, which Android malicious application variants emerging in endlessly and growth rapidly, this paper is based on machine learning algorithms and focuses on the research of malicious application detection algorithms under the Android platform. Based on that we proposes a lightweight Android malware detection and identification algorithm. Aiming at the requirement of lightweight model, In response to the needs of lightweight models, the feature selection method based on support filtering and Lasso LR model is adopted to greatly reduce the feature space. Combining the characteristics of high feature dimension, using the field-aware decomposition machine (FFM) model as the classifier, the detection performance with an F1 value of 0.990887 is achieved, and the accuracy of the detection of malicious applications is improved.
{"title":"Android malicious application detection based on Support Filtering and Lasso LR Algorithm","authors":"Le Weng, Hengyu Liu, Lianfeng Huang, Yingmin Zhang, Chao Feng","doi":"10.1109/CCCI52664.2021.9583199","DOIUrl":"https://doi.org/10.1109/CCCI52664.2021.9583199","url":null,"abstract":"With the vigorous development of the mobile Internet, the Android system, which accounts for 76% of the drama mobile operating system, has also been widely promoted and popularized. However, due to its own open source characteristics, various parts of the Android system are facing serious threats from hacker attacks, and the main threat comes from malicious applications. In order to cope with the challenge, which Android malicious application variants emerging in endlessly and growth rapidly, this paper is based on machine learning algorithms and focuses on the research of malicious application detection algorithms under the Android platform. Based on that we proposes a lightweight Android malware detection and identification algorithm. Aiming at the requirement of lightweight model, In response to the needs of lightweight models, the feature selection method based on support filtering and Lasso LR model is adopted to greatly reduce the feature space. Combining the characteristics of high feature dimension, using the field-aware decomposition machine (FFM) model as the classifier, the detection performance with an F1 value of 0.990887 is achieved, and the accuracy of the detection of malicious applications is improved.","PeriodicalId":136382,"journal":{"name":"2021 International Conference on Communications, Computing, Cybersecurity, and Informatics (CCCI)","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127341856","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 : 2021-03-16DOI: 10.1109/CCCI52664.2021.9583209
Ashish Rana, A. Malhi
The highway-env reinforcement learning tasks provides a good abstract testbed for designing driving agents for specific driving scenarios like lane changing, parking or intersections etc. But, generally these driving simulation environments often restrict themselves to safer and precise trajectories. However, we clearly know that real driving tasks often involve very high risk collision prone unexpected situations. Hence, the autonomous model-free driving agents prepared in these environments are blind to certain low probability traffic collision corner cases. In our study we systematically focus on generating adversarial driving collision prone scenarios with dangerous driving behavior and heavy traffic in order to create robust autonomous agents. In our experimentation we train model free learning agents with additional collision prone scenario simulations and compare their efficacy with regular simulation based agents. Ultimately, we create a causal experimentation setup which successfully accounts for the performance improvements across different driving scenarios by utilizing learning from risky driving situations.
{"title":"Building Safer Autonomous Agents by Leveraging Risky Driving Behavior Knowledge","authors":"Ashish Rana, A. Malhi","doi":"10.1109/CCCI52664.2021.9583209","DOIUrl":"https://doi.org/10.1109/CCCI52664.2021.9583209","url":null,"abstract":"The highway-env reinforcement learning tasks provides a good abstract testbed for designing driving agents for specific driving scenarios like lane changing, parking or intersections etc. But, generally these driving simulation environments often restrict themselves to safer and precise trajectories. However, we clearly know that real driving tasks often involve very high risk collision prone unexpected situations. Hence, the autonomous model-free driving agents prepared in these environments are blind to certain low probability traffic collision corner cases. In our study we systematically focus on generating adversarial driving collision prone scenarios with dangerous driving behavior and heavy traffic in order to create robust autonomous agents. In our experimentation we train model free learning agents with additional collision prone scenario simulations and compare their efficacy with regular simulation based agents. Ultimately, we create a causal experimentation setup which successfully accounts for the performance improvements across different driving scenarios by utilizing learning from risky driving situations.","PeriodicalId":136382,"journal":{"name":"2021 International Conference on Communications, Computing, Cybersecurity, and Informatics (CCCI)","volume":"109 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131148180","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：