More Like this

1. Anytime Approximate Bi-Objective Search

   Zhang, H.; Salzman, O.; Kumar, S.; Felner, A.; Hernandez, C.; Koenig, S. (January 2022, Proceedings of the Symposium on Combinatorial Search (SoCS))

   The Pareto-optimal frontier for a bi-objective search problem instance consists of all solutions that are not worse than any other solution in both objectives. The size of the Pareto-optimal frontier can be exponential in the size of the input graph, and hence finding it can be hard. Some existing works leverage a user-specified approximation factor ε to compute an approximate Pareto-optimal frontier that can be significantly smaller than the Pareto-optimal frontier. In this paper, we propose an anytime approximate bi-objective search algorithm, called Anytime Bi-Objective A*-ε (A-BOA*ε). A-BOA*ε is useful when deliberation time is limited. It first finds an approximate Pareto-optimal frontier quickly, iteratively improves it while time allows, and eventually finds the Pareto-optimal frontier. It efficiently reuses the search effort from previous iterations and makes use of a novel pruning technique. Our experimental results show that A-BOA*ε substantially outperforms baseline algorithms that do not reuse previous search effort, both in terms of runtime and number of node expansions. In fact, the most advanced variant of A-BOA*ε even slightly outperforms BOA*, a state-of-the-art bi-objective search algorithm, for finding the Pareto-optimal frontier. Moreover, given only a limited amount of deliberation time, A-BOA*ε finds solutions that collectively approximate the Pareto-optimal frontier much better than the solutions found by BOA*. 

   more » « less
2. Balancing Competing Objectives with Noisy Data: Score-Based Classifiers for Welfare-Aware Machine Learning

   Rolf, Esther; Simchowitz, Max; Dean, Sarah; Liu, Lydia T.; Bjorkegren, Daniel; Hardt, Moritz; Blumenstock, Joshua (April 2020, Proceedings of the 37th International Conference on Machine Learning)

   null (Ed.)

   While real-world decisions involve many competing objectives, algorithmic decisions are often evaluated with a single objective function. In this paper, we study algorithmic policies which explicitly trade off between a private objective (such as profit) and a public objective (such as social welfare). We analyze a natural class of policies which trace an empirical Pareto frontier based on learned scores, and focus on how such decisions can be made in noisy or data-limited regimes. Our theoretical results characterize the optimal strategies in this class, bound the Pareto errors due to inaccuracies in the scores, and show an equivalence between optimal strategies and a rich class of fairness-constrained profit-maximizing policies. We then present empirical results in two different contexts — online content recommendation and sustainable abalone fisheries — to underscore the generality of our approach to a wide range of practical decisions. Taken together, these results shed light on inherent trade-offs in using machine learning for decisions that impact social welfare. 

   more » « less
3. Balancing Competing Objectives with Noisy Data: Score-Based Classifiers for Welfare-Aware Machine Learning

   Rolf, Esther; Simchowitz, Max; Dean, Sarah; Liu, Lydia; Bjorkegren, Daniel; Hardt, Moritz; Blumenstock, Joshua E (January 2020, Proceedings of Machine Learning Research)

   null (Ed.)

   While real-world decisions involve many competing objectives, algorithmic decisions are often evaluated with a single objective function. In this paper, we study algorithmic policies which explicitly trade off between a private objective (such as profit) and a public objective (such as social welfare). We analyze a natural class of policies which trace an empirical Pareto frontier based on learned scores, and focus on how such decisions can be made in noisy or data-limited regimes. Our theoretical results characterize the optimal strategies in this class, bound the Pareto errors due to inaccuracies in the scores, and show an equivalence between optimal strategies and a rich class of fairness-constrained profit-maximizing policies. We then present empirical results in two different contexts — online content recommendation and sustainable abalone fisheries — to underscore the generality of our approach to a wide range of practical decisions. Taken together, these results shed light on inherent trade-offs in using machine learning for decisions that impact social welfare. 

   more » « less
4. Multi-objective Privacy-preserving Text Representation Learning

   https://doi.org/10.1145/3459637.3482147  

   Zhan, Huixin; Zhang, Kun; Hu, Chenyi; Sheng, Victor (October 2021, Proceedings of the 30th ACM International Conference on Information & Knowledge Management)

   are explicitly contained in the text or be implicit. For example, demographic information about the author of a text can be predicted with above-chance accuracy from linguistic cues in the text itself. Letting alone its explicitness, some of the private information correlates with the output labels and therefore can be learned by a neural network. In such a case, there is a tradeoff between the utility of the representation (measured by the accuracy of the classification network) and its privacy. This problem is inherently a multi-objective problem because these two objectives may conflict, necessitating a trade-off. Thus, we explicitly cast this problem as multi-objective optimization (MOO) with the overall objective of finding a Pareto stationary solution. We, therefore, propose a multiple-gradient descent algorithm (MGDA) that enables the efficient application of the Frank-Wolfe algorithm [10] using the line search. Experimental results on sentiment analysis and part-of-speech (POS) tagging show that MGDA produces higher-performing models than most recent proxy objective approaches, and performs as well as single objective baselines. 

   more » « less
5. Certification and Trade-off of Multiple Fairness Criteria in Graph-based Spam Detection

   https://doi.org/10.1145/3459637.3482325  

   Burkholder, Kai; Kwock, Kenny; Xu, Yuesheng; Liu, Jiaxin; Chen, Chao; Xie, Sihong (October 2021, Proceedings of the 30th ACM International Conference on Information & Knowledge Management)

   Spamming reviews are prevalent in review systems to manipulate seller reputation and mislead customers. Spam detectors based on graph neural networks (GNN) exploit representation learning and graph patterns to achieve state-of-the-art detection accuracy. The detection can influence a large number of real-world entities and it is ethical to treat different groups of entities as equally as possible. However, due to skewed distributions of the graphs, GNN can fail to meet diverse fairness criteria designed for different parties. We formulate linear systems of the input features and the adjacency matrix of the review graphs for the certification of multiple fairness criteria. When the criteria are competing, we relax the certification and design a multi-objective optimization (MOO) algorithm to explore multiple efficient trade-offs, so that no objective can be improved without harming another objective. We prove that the algorithm converges to a Pareto efficient solution using duality and the implicit function theorem. Since there can be exponentially many trade-offs of the criteria, we propose a data-driven stochastic search algorithm to approximate Pareto fronts consisting of multiple efficient trade-offs. Experimentally, we show that the algorithms converge to solutions that dominate baselines based on fairness regularization and adversarial training. 

   more » « less