Ming Yi

Xiaolei Liu, Ming Yi, Kangyi Ding et al.

Deep learning models are vulnerable to backdoor attacks, where attackers inject malicious behavior through data poisoning and later exploit triggers to manipulate deployed models. To improve the stealth and effectiveness of backdoors, prior studies have introduced various imperceptible attack methods targeting both defense mechanisms and manual inspection. However, all poisoning-based attacks still rely on privileged access to the training dataset. Consequently, model distillation using a trusted dataset has emerged as an effective defense against these attacks. To bridge this gap, we introduce INK, an inheritable natural backdoor attack that targets model distillation. The key insight behind INK is the use of naturally occurring statistical features in all datasets, allowing attackers to leverage them as backdoor triggers without direct access to the training data. Specifically, INK employs image variance as a backdoor trigger and enables both clean-image and clean-label attacks by manipulating the labels and image variance in an unauthenticated dataset. Once the backdoor is embedded, it transfers from the teacher model to the student model, even when defenders use a trusted dataset for distillation. Theoretical analysis and experimental results demonstrate the robustness of INK against transformation-based, search-based, and distillation-based defenses. For instance, INK maintains an attack success rate of over 98\% post-distillation, compared to an average success rate of 1.4\% for existing methods.

Ming Yi, Meng Wang

The rapid deployment of renewable generations such as photovoltaic (PV) generations brings great challenges to the resiliency of existing power systems. Because PV generations are volatile and typically invisible to the power system operator, estimating the generation and characterizing the uncertainty are in urgent need for operators to make insightful decisions. This paper summarizes our recent results on energy disaggregation at the substation level with Behind-the-Meter solar generation. We formulate the so-called ``partial label'' problem for energy disaggregation at substations, where the aggregate measurements contain the total consumption of multiple loads, and the existence of some loads is unknown. We develop two model-free disaggregation approaches based on deterministic dictionary learning and Bayesian dictionary learning, respectively. Unlike conventional methods which require fully annotated training data of individual loads, our approaches can extract load patterns given partially labeled aggregate data. Therefore, our partial label formulation is more applicable in the real world. Compared with deterministic dictionary learning, the Bayesian dictionary learning-based approach provides the uncertainty measure for the disaggregation results, at the cost of increased computational complexity. All the methods are validated by numerical experiments.

Saud Alghumayjan, Jiajun Han, Ningkun Zheng et al.

This paper presents an integrated model for bidding energy storage in day-ahead and real-time markets to maximize profits. We show that in integrated two-stage bidding, the real-time bids are independent of day-ahead settlements, while the day-ahead bids should be based on predicted real-time prices. We utilize a transformer-based model for real-time price prediction, which captures complex dynamical patterns of real-time prices, and use the result for day-ahead bidding design. For real-time bidding, we utilize a long short-term memory-dynamic programming hybrid real-time bidding model. We train and test our model with historical data from New York State, and our results showed that the integrated system achieved promising results of almost a 20\% increase in profit compared to only bidding in real-time markets, and at the same time reducing the risk in terms of the number of days with negative profits.