More Like this

1. Symbolic Metamodels for Interpreting Black-Boxes Using Primitive Functions

   https://doi.org/10.1609/aaai.v37i6.25816  

   Abroshan, Mahed; Mishra, Saumitra; Khalili, Mohammad Mahdi (June 2023, Proceedings of the AAAI Conference on Artificial Intelligence)

   One approach for interpreting black-box machine learning models is to find a global approximation of the model using simple interpretable functions, which is called a metamodel (a model of the model). Approximating the black-box witha metamodel can be used to 1) estimate instance-wise feature importance; 2) understand the functional form of the model; 3) analyze feature interactions. In this work, we propose a new method for finding interpretable metamodels. Our approach utilizes Kolmogorov superposition theorem, which expresses multivariate functions as a composition of univariate functions (our primitive parameterizedfunctions). This composition can be represented in the form of a tree. Inspired by symbolic regression, we use a modified form of genetic programming to search over different tree configurations. Gradient descent (GD) is used to optimize the parameters of a given configuration. Our method is a novel memetic algorithm that uses GD not only for training numerical constants but also for the trainingof building blocks. Using several experiments, we show that our method outperforms recent metamodeling approaches suggested for interpreting black-boxes. 

   more » « less
2. Seismic Ocean Thermometry Using CTBTO Hydrophones

   https://doi.org/10.1029/2023JB026687  

   Wu, Wenbo; Shen, Zhichao; Peng, Shirui; Zhan, Zhongwen; Callies, Jörn (September 2023, Journal of Geophysical Research: Solid Earth)

   Abstract Due to limited observational coverage, monitoring the warming of the global ocean, especially the deep ocean, remains a challenging sampling problem. Seismic ocean thermometry (SOT) complements existing point measurements by inferring large‐scale averaged ocean temperature changes using the sound waves generated by submarine earthquakes, calledTwaves. We demonstrate here that Comprehensive Nuclear‐Test‐Ban Treaty Organization (CTBTO) hydrophones can recordTwaves with a higher signal‐to‐noise ratio compared to a previously used land‐basedT‐wave station. This allows us to use small earthquakes (magnitude <4.0), which occur much more frequently than large events, dramatically improving the resulting temporal resolution of SOT. We also find that the travel time changes ofTwaves at the land‐basedT‐wave station and the CTBTO hydrophone show small but systematic differences, although the two stations are only about 20 km apart. We attribute this feature to their different acoustic mode components sampling different parts of the ocean. Applying SOT to two CTBTO hydrophones in the East Indian Ocean reveals signals from decadal warming, seasonal variations, and mesoscale eddies, some of which are missing or underestimated in previously available temperature reconstructions. This application demonstrates the great advantage of hydrophone stations for global SOT, especially in regions with a low seismicity level. 

   more » « less
3. Principled Reinforcement Learning with Human Feedback from Pairwise or K-wise Comparisons

   Zhu, Banghua and (January 2023, Proceedings of Machine Learning Research)

   Krause, Andreas and (Ed.)

   We provide a theoretical framework for Reinforcement Learning with Human Feedback (RLHF). We show that when the underlying true reward is linear, under both Bradley-Terry-Luce (BTL) model (pairwise comparison) and Plackett-Luce (PL) model ($$K$$-wise comparison), MLE converges under certain semi-norm for the family of linear reward. On the other hand, when training a policy based on the learned reward model, we show that MLE fails while a pessimistic MLE provides policies with good performance under certain coverage assumption. We also show that under the PL model, both the true MLE and a different MLE which splits the $$K$$-wise comparison into pairwise comparisons converge, while the true MLE is asymptotically more efficient. Our results validate the empirical success of the existing RLHF algorithms, and provide new insights for algorithm design. Our analysis can also be applied for the problem of online RLHF and inverse reinforcement learning. 

   more » « less
4. An Efficient Training Algorithm for Models with Block-wise Sparsity

   Zhu, Ding; Zuo, Zhiqun; Khalili, Mohammad_Mahdi (March 2025, Transactions on machine learning research)

   Large-scale machine learning (ML) models are increasingly being used in critical domains like education, lending, recruitment, healthcare, criminal justice, etc. However, the training, deployment, and utilization of these models demand substantial computational resources. To decrease computation and memory costs, machine learning models with sparse weight matrices are widely used in the literature. Among sparse models, those with special sparse structures (e.g., models with block-wise sparse weight matrices) fit better with the hardware accelerators and can decrease the memory and computation costs during the inference. Unfortunately, while there are several efficient training methods, none of them are designed to train a block-wise sparse model efficiently. As a result, the current methods for training block-wise sparse models start with full and dense models leading to inefficient training. In this work, we focus on training models with block-wise sparse matrices and propose an efficient training algorithm to decrease both computation and memory costs during training and inference. In addition, we will show that our proposed method enables us to efficiently find the right block size for the sparsity pattern during the training process. Our extensive empirical and theoretical analyses show that our algorithms can decrease the computation and memory costs significantly without a performance drop compared to baselines. 

   more » « less
5. Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation

   Tseng, Hung-Yu; Lee, Hsin-Ying; Huang, Jia-Bin; Yang, Ming-Hsuan (May 2020, International Conference on Learning Representations)

   Few-shot classification aims to recognize novel categories with only few labeled images in each class. Existing metric-based few-shot classification algorithms predict categories by comparing the feature embeddings of query images with those from a few labeled images (support examples) using a learned metric function. While promising performance has been demonstrated, these methods often fail to generalize to unseen domains due to large discrepancy of the feature distribution across domains. In this work, we address the problem of few-shot classification under domain shifts for metric-based methods. Our core idea is to use feature-wise transformation layers for augmenting the image features using affine transforms to simulate various feature distributions under different domains in the training stage. To capture variations of the feature distributions under different domains, we further apply a learning-to-learn approach to search for the hyper-parameters of the feature-wise transformation layers. We conduct extensive experiments and ablation studies under the domain generalization setting using five few-shot classification datasets: mini-ImageNet, CUB, Cars, Places, and Plantae. Experimental results demonstrate that the proposed feature-wise transformation layer is applicable to various metric-based models, and provides consistent improvements on the few-shot classification performance under domain shift. 

   more » « less