IoT 23(1): e5

Research Article

Robustness of Classification Algorithm in the Face of Label Noise

Download168 downloads
  • @ARTICLE{10.4108/eetiot.v9i1.3270,
        author={Jiawei ZHAO and Mengyao KANG and Zheng HAN},
        title={Robustness of Classification Algorithm in the Face of Label Noise},
        journal={EAI Endorsed Transactions on Internet of Things},
        volume={9},
        number={1},
        publisher={EAI},
        journal_a={IOT},
        year={2023},
        month={6},
        keywords={Label noise, Machine learning, Transition matrix, Robustness of algorithm},
        doi={10.4108/eetiot.v9i1.3270}
    }
    
  • Jiawei ZHAO
    Mengyao KANG
    Zheng HAN
    Year: 2023
    Robustness of Classification Algorithm in the Face of Label Noise
    IOT
    EAI
    DOI: 10.4108/eetiot.v9i1.3270
Jiawei ZHAO1,*, Mengyao KANG1, Zheng HAN1
  • 1: University of Sydney
*Contact email: zjweiok@gmail.com

Abstract

Label noise is an important part in the process of machine learning. Transition matrix provides an effective way to reduce the impact of label noise on classification algorithm. In this experiment, we study logistic regression algorithm and random forest algorithm. We use the known real transition matrix to evaluate the robustness of the algorithm on two datasets. We also design a transition matrix estimator to estimate the transition matrix of three datasets and evaluate the robustness of the two algorithms. We use average error to evaluate the effectiveness of the transition matrix estimator and the top-1 accuracy to evaluate our method.