Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Data Valuation using Reinforcement Learning

About

Quantifying the value of data is a fundamental problem in machine learning. Data valuation has multiple important use cases: (1) building insights about the learning task, (2) domain adaptation, (3) corrupted sample discovery, and (4) robust learning. To adaptively learn data values jointly with the target task predictor model, we propose a meta learning framework which we name Data Valuation using Reinforcement Learning (DVRL). We employ a data value estimator (modeled by a deep neural network) to learn how likely each datum is used in training of the predictor model. We train the data value estimator using a reinforcement signal of the reward obtained on a small validation set that reflects performance on the target task. We demonstrate that DVRL yields superior data value estimates compared to alternative methods across different types of datasets and in a diverse set of application scenarios. The corrupted sample discovery performance of DVRL is close to optimal in many regimes (i.e. as if the noisy samples were known apriori), and for domain adaptation and robust learning DVRL significantly outperforms state-of-the-art by 14.6% and 10.8%, respectively.

Jinsung Yoon, Sercan O. Arik, Tomas Pfister• 2019

Related benchmarks

TaskDatasetResultRank
ClassificationDIGITS (test)
Average Accuracy42.2
59
ClassificationElectricity (test)
Accuracy63.7
55
Classificationfried (test)
Mean Test Accuracy73
10
Classificationelection (test)
Mean Test Accuracy56.5
10
RegressionGaussian 1K seller (train)
MSE1.33
10
RegressionMIMIC 1K seller (train)
MSE229.7
10
Classification2dplanes (test)
Mean Test Accuracy72.8
10
ClassificationMINIBOONE (test)
Mean Test Accuracy68.9
10
Classificationnomao (test)
Mean Test Accuracy77.7
10
Classificationbbc-embeddings (test)
Mean Test Accuracy65.3
10
Showing 10 of 15 rows

Other info

Follow for update