Data valuation has emerged as a powerful framework for quantifying each datum's

However, a validation set may not always be available in practice [35]. Also, as different choices ofthe validation set can lead todifferent data valuations, itischallenging for the data providers to agree on the choice ofsuch avalidation set [35].

Data valuation is a class of techniques for quantitatively assessing the value of data for applications like pricing in data marketplaces.

The critical challenge when using the Shapley value is its well-known computational intractability.

Data valuation, a growing field that aims at quantifying the usefulness of individual data sources for training machine learning (ML) models, faces notable yet often overlooked privacy challenges.

Data valuation, or the valuation of individual datum contributions, has seen growing interest in machine learning due to its demonstrable efficacy for tasks such as noisy label detection. In particular, due to the desirable axiomatic properties, several Shapley value approximations have been proposed. In these methods, the value function is usually defined as the predictive accuracy over the entire development set. However, this limits the ability to differentiate between training instances that are helpful or harmful to their own classes. Intuitively, instances that harm their own classes may be noisy or mislabeled and should receive a lower valuation than helpful instances. In this work, we propose CS-Shapley, a Shapley value with a new value function that discriminates between training instances' in-class and out-of-class contributions. Our theoretical analysis shows the proposed value function is (essentially) the unique function that satisfies two desirable properties for evaluating data values in classification. Further, our experiments on two benchmark evaluation tasks (data removal and noisy label detection) and four classifiers demonstrate the effectiveness of CS-Shapley over existing methods. Lastly, we evaluate the "transferability" of data values estimated from one classifier to others, and our results suggest Shapley-based data valuation is transferable for application across different models.