A Meta-Learning Approach for Custom Model Training

被引:0
|
作者
Eshratifar, Amir Erfan [1 ]
Abrishami, Mohammad Saeed [1 ]
Eigen, David [2 ]
Pedram, Massoud [1 ]
机构
[1] Univ Southern Calif, Dept Elect Engn, Los Angeles, CA 90089 USA
[2] Clarifai, San Francisco, CA 94105 USA
来源
THIRTY-THIRD AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FIRST INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE / NINTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE | 2019年
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Transfer-learning and meta-learning are two effective methods to apply knowledge learned from large data sources to new tasks. In few-class, few-shot target task settings (i.e. when there are only a few classes and training examples available in the target task), meta-learning approaches that optimize for future task learning have outperformed the typical transfer approach of initializing model weights from a pretrained starting point. But as we experimentally show, meta-learning algorithms that work well in the few-class setting do not generalize well in many-shot and many-class cases. In this paper, we propose a joint training approach that combines both transfer-learning and meta-learning. Benefiting from the advantages of each, our method obtains improved generalization performance on unseen target tasks in both few- and many-class and few- and many-shot scenarios.
引用
收藏
页码:9937 / 9938
页数:2
相关论文
共 50 条
  • [1] Variational HyperAdam: A Meta-Learning Approach to Network Training
    Wang, Shipeng
    Yang, Yan
    Sun, Jian
    Xu, Zongben
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2022, 44 (08) : 4469 - 4484
  • [2] A Meta-Learning Approach for Training Explainable Graph Neural Networks
    Spinelli, Indro
    Scardapane, Simone
    Uncini, Aurelio
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2024, 35 (04) : 4647 - 4655
  • [3] SPEAKER ADAPTIVE TRAINING USING MODEL AGNOSTIC META-LEARNING
    Klejch, Ondrej
    Fainberg, Joachim
    Bell, Peter
    Renals, Steve
    2019 IEEE AUTOMATIC SPEECH RECOGNITION AND UNDERSTANDING WORKSHOP (ASRU 2019), 2019, : 881 - 888
  • [4] Learning to Defer to a Population: A Meta-Learning Approach
    Tailor, Dharmesh
    Patra, Aditya
    Verma, Rajeev
    Manggala, Putra
    Nalisnick, Eric
    INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 238, 2024, 238
  • [5] Meta-Learning to Improve Pre-Training
    Raghu, Aniruddh
    Lorraine, Jonathan
    Kornblith, Simon
    McDermott, Matthew
    Duvenaud, David
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 34 (NEURIPS 2021), 2021, 34
  • [6] A Meta-Learning Approach to Error Prediction
    Guimaraes, Miguel
    Carneiro, Davide
    PROCEEDINGS OF 2021 16TH IBERIAN CONFERENCE ON INFORMATION SYSTEMS AND TECHNOLOGIES (CISTI'2021), 2021,
  • [7] A Meta-learning Approach to Fair Ranking
    Wang, Yuan
    Tao, Zhiqiang
    Fang, Yi
    PROCEEDINGS OF THE 45TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL (SIGIR '22), 2022, : 2539 - 2544
  • [8] Improved Meta-learning Training for Speaker Verification
    Chen, Yafeng
    Guo, Wu
    Gu, Bin
    INTERSPEECH 2021, 2021, : 1049 - 1053
  • [9] Meta-learning: searching in the model space
    Duch, W
    Grudzinski, K
    8TH INTERNATIONAL CONFERENCE ON NEURAL INFORMATION PROCESSING, VOLS 1-3, PROCEEDING, 2001, : 235 - 240
  • [10] Learning to adapt: a meta-learning approach for speaker adaptation
    Klejch, Ondrej
    Fainberg, Joachim
    Bell, Peter
    19TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2018), VOLS 1-6: SPEECH RESEARCH FOR EMERGING MARKETS IN MULTILINGUAL SOCIETIES, 2018, : 867 - 871