TRACE: Theoretical Risk Attribution under Covariate-shift Effects

When a source-trained model $Q$ is replaced by a model $\tilde{Q}$ trained on shifted data, its performance on the source domain can change unpredictably. To address this, we study the two-model risk change, $ΔR := R_P(\tilde{Q}) - R_P(Q)$, under covariate shift. We introduce TRACE (Theoretical Risk Attribution under Covariate-shift Effects), a framework that decomposes $|ΔR|$ into an interpretable upper bound. This decomposition disentangles the risk change into four actionable factors: two generalization gaps, a model change penalty, and a covariate shift penalty, transforming the bound into a powerful diagnostic tool for understanding why performance has changed. To make TRACE a fully computable diagnostic, we instantiate each term. The covariate shift penalty is estimated via a model sensitivity factor (from high-quantile input gradients) and a data-shift measure; we use feature-space Optimal Transport (OT) by default and provide a robust alternative using Maximum Mean Discrepancy (MMD). The model change penalty is controlled by the average output distance between the two models on the target sample. Generalization gaps are estimated on held-out data. We validate our framework in an idealized linear regression setting, showing the TRACE bound correctly captures the scaling of the true risk difference with the magnitude of the shift. Across synthetic and vision benchmarks, TRACE diagnostics are valid and maintain a strong monotonic relationship with the true performance degradation. Crucially, we derive a deployment gate score that correlates strongly with $|ΔR|$ and achieves high AUROC/AUPRC for gating decisions, enabling safe, label-efficient model replacement.

Fundamental Trade-Offs in Monostatic ISAC: A Holistic Investigation Towards 6G

Integrated sensing and communication (ISAC) emerges as a cornerstone technology for the upcoming 6G era, seamlessly incorporating sensing functionality into wireless networks as an inherent capability. This paper undertakes a holistic investigation of two fundamental trade-offs in monostatic OFDM ISAC systems-namely, the time-frequency domain trade-off and the spatial domain trade-off. To ensure robust sensing across diverse modulation orders in the time-frequency domain, including high-order QAM, we design a linear minimum mean-squared-error (LMMSE) estimator tailored for sensing with known, randomly generated signals of varying amplitude. Moreover, we explore spatial domain trade-offs through two ISAC transmission strategies: concurrent, employing joint beams, and time-sharing, using separate, time-non-overlapping beams for sensing and communications. Simulations demonstrate superior performance of the LMMSE estimator in detecting weak targets in the presence of strong ones under high-order QAM, consistently yielding more favorable ISAC trade-offs than existing baselines. Key insights into these trade-offs under various modulation schemes, SNR conditions, target radar cross section (RCS) levels and transmission strategies highlight the merits of the proposed LMMSE approach.

Adjustable Privacy using Autoencoder-based Learning Structure

Inference centers need more data to have a more comprehensive and beneficial learning model, and for this purpose, they need to collect data from data providers. On the other hand, data providers are cautious about delivering their datasets to inference centers in terms of privacy considerations. In this paper, by modifying the structure of the autoencoder, we present a method that manages the utility-privacy trade-off well. To be more precise, the data is first compressed using the encoder, then confidential and non-confidential features are separated and uncorrelated using the classifier. The confidential feature is appropriately combined with noise, and the non-confidential feature is enhanced, and at the end, data with the original data format is produced by the decoder. The proposed architecture also allows data providers to set the level of privacy required for confidential features. The proposed method has been examined for both image and categorical databases, and the results show a significant performance improvement compared to previous methods.