陈俊洁-深度学习系统的性能提升.pdf

《陈俊洁-深度学习系统的性能提升.pdf》由会员分享，可在线阅读，更多相关《陈俊洁-深度学习系统的性能提升.pdf（44页珍藏版）》请在三个皮匠报告上搜索。

1、深度学习系统的性能提升陈俊洁 天津大学演讲嘉宾陈俊洁国家优青，天津大学特聘研究员，博导，软件工程团队负责人研究方向主要为基础软件测试、可信人工智能、数据驱动的软件工程等。荣获中国科协青年托举人才、CCF优博、电子学会自然科学一等奖等奖项。近年共发表学术论文70篇，其中CCF A类论文50余篇，获六项最佳论文奖（包括五项CCF-A类会议ACM SIGSOFT Distinguished Paper Award，以及一项CCF-B类会议ISSRE的Best Research Paper Award）。成果在华为、百度等多家知名企业落地。担任CCF-A类会议ASE 2021评审过程主席，Dagstu

2、hl研讨会联合主席，以及软件工程领域全部CCF-A类会议的程序委员会成员等。目 录CONTENTS1.深度学习系统的回归性能提升2.深度代码模型的鲁棒性能力提升3.深度代码模型部署后性能即时提升深度学习系统的回归性能提升PART 01Regression in Deep Learning SystemsIt is important to detect regression faults!DL System Ver1.0DL System Ver2.0DL System Ver3.0New RequirementsFixing/ImprovementAccuracy:40%Accuracy:6

3、0%Existing Works Have LimitationsSOTA techniques can not be directly adapt to solve this issue.DL System Ver1.0Accuracy=91%DL System Ver2.0Accuracy=91.5%Code ChangeNeuron ChangeRegression Fuzzing in Traditional SoftwareDL Systems do not have explicit logical structuresNeuron change nearly affect all

4、 the neurons while code change only affect limited partsFuzzing for Deep Learning ModelsIgnore the difference between different versions of the DL modelsOverlook important properties of the testing,such as fidelity and diversity.DeepHunter:Fuzzing guided by fine-grained neuron coverage in a specific

5、 version DiffChaser:Detect disagreements in Quantization by generating test cases toward decision boundary locates code changes in software evolution and utilize them to guide the regression fuzzingIgnores Difference:Poor Fault-TriggeringOverlooksFidelity&Diversity1122Our Idea of DRFuzzChallenge 1:F

6、ault-TriggeringSolution:Amplifying the prediction difference between versions through effective mutation to trigger more faults.Challenge 3:DiversitySolution:Using seed maintenance to generate test inputs trigger different regression faults.Challenge 2:FidelitySolution:Designing GAN-based fidelity a