Influence Selection for Active Learning

Liu, Zhuoming; Ding, Hao; Zhong, Huaping; Li, Weijia; Dai, Jifeng; He, Conghui

doi:10.1109/iccv48922.2021.00914

Cited by 36 publications

(9 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Taking our method as an example, we use the classifier trained on the training set to infer unlabeled sample data, and then select unlabeled data samples as candidate samples to label information through the inference results. However, whether the classifier in this process is biased, it biases the categories that have been recognized skillfully and ignores the samples that have not been skillfully classified in the sample selection stage 32 . The root cause of the data bias problem is the preference for partial data in the sampling mechanism itself.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Efficient data-centric pest images identification method based on Mahalanobis entropy for intelligent agriculture

Zhu

2023

J. Electron. Imag.

View full text Add to dashboard Cite

The frequent outbreak of crop pests is one of the main factors affecting crop yield.The accurate identification of pests is significant for effective pest control, which is the basis for the safe growth of crops and is also an important guarantee for high crop quality and crop yield. Computer vision based on supervised deep learning enables intelligent identification of pests, whose success is still inseparable from a large amount of labeled data, causing a large number of resource consumption due to data labeling. Therefore, an in-depth study on maximizing the value of data is essential when lacking labeled data. A new data evaluation method based on Mahalanobis distance and entropy is proposed to address the problem of lacking labeled data in intelligent pest identification. This method enables filtering high-value data, thus achieving effective pest identification performance with a small data volume. The experiment is conducted on a dataset we collected called PD-20, which shows the proposed method achieves baseline accuracy of 100% using only about 60% of the original data. Moreover, the proposed method can save at least 10% of the data volume compared with three comparison methods. To facilitate the deep integration of smart agriculture with artificial intelligence (AI), we designed an interactive framework of active learning for pest identification based on the proposed method, which lays the foundation for the application of AI in direction of agriculture.

show abstract

Section: Discussionmentioning

confidence: 99%

“…However, whether the classifier in this process is biased, it biases the categories that have been recognized skillfully and ignores the samples that have not been skillfully classified in the sample selection stage. 32 The root cause of the data bias problem is the preference for partial data in the sampling mechanism itself.…”

Section: Sbd Mechanismmentioning

confidence: 99%

Efficient data-centric pest images identification method based on Mahalanobis entropy for intelligent agriculture

Zhu

2023

J. Electron. Imag.

View full text Add to dashboard Cite

show abstract

“…To improve the sampling bias of any baseline sampling method Settles and Craven propose an information density method which is computationally expensive for large pools of unlabeled data [4]. Several other successful methods have been proposed to combat the sampling bias and robustness issues, but all these methods are designed to work with specific model types, e.g., with convolutional neural networks [12], [13], [15], [16], [28]. A generalizable method for reducing sampling bias is proposed by Elhamifar et al [17], but it involves an optimization problem with n 2 variables where n is the number of datapoints in the data set, so the method is not scalable to large data sets.…”

Section: A Active Learningmentioning

confidence: 99%

Active Learning with Combinatorial Coverage

Prathyush¹,

Cody²,

Beling³

et al. 2023

Preprint

View full text Add to dashboard Cite

Active learning is a practical field of machine learning that automates the process of selecting which data to label. Current methods are effective in reducing the burden of data labeling but are heavily model-reliant. This has led to the inability of sampled data to be transferred to new models as well as issues with sampling bias. Both issues are of crucial concern in machine learning deployment. We propose active learning methods utilizing combinatorial coverage to overcome these issues. The proposed methods are data-centric, as opposed to modelcentric, and through our experiments we show that the inclusion of coverage in active learning leads to sampling data that tends to be the best in transferring to better performing models and has a competitive sampling bias compared to benchmark methods.

show abstract

“…Both of these methods are only applicable to convolutional neural networks. Liu et al propose a influence based selection method, but the data point that has most influence on the gradient of the models' loss function may not necessarily improve the performance of the model the most, and could be an outlier [21].…”

Section: Active Learning Literature Reviewmentioning

confidence: 99%

Active Learning with Combinatorial Coverage

Prathyush

Cody

Beling

et al. 2022

2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)

View full text Add to dashboard Cite

Active learning is a practical field of machine learning as labeling data or determining which data to label can be a time consuming and inefficient task. Active learning automates the process of selecting which data to label, but current methods are heavily model reliant. This has led to the inability of sampled data to be transferred to new models as well as issues with sampling bias. Both issues are of crucial concern in machine learning deployment. We propose active learning methods utilizing Combinatorial Coverage to overcome these issues.The proposed methods are data-centric, and through our experiments we show that the inclusion of coverage in active learning leads to sampling data that tends to be the best in transferring to different models and has a competitive sampling bias compared to benchmark methods. been possible. I would like to extend my sincerest gratitude to Dr. Laura Freeman, who gave me the opportunity, support, and valuable input necessary to work on this research. I would also like to thank Dr. Manish Bansal for the constructive feedback and support in both the classroom and research work. Without the aid of all the professors in my masters degree committee I would not have finished this work. Last but not least, I would like to thank my parents and brother; Aruna, Shanker, and Aathish, for their support in education

show abstract

Influence Selection for Active Learning

Cited by 36 publications

References 20 publications

Efficient data-centric pest images identification method based on Mahalanobis entropy for intelligent agriculture

Efficient data-centric pest images identification method based on Mahalanobis entropy for intelligent agriculture

Active Learning with Combinatorial Coverage

Active Learning with Combinatorial Coverage

Contact Info

Product

Resources

About