arXiv subjects

David Love

Publications and source records attributed to David Love.

7 recordsLinked to original sources

arXiv2026

SHIELD: A Diverse Clinical Note Dataset and Distilled Small Language Models for Enterprise-Scale De-identification

De-identification of clinical text is a prerequisite for the secondary use of electronic health records. Existing public benchmarks such as the i2b2 2006 and 2014 corpora are over a decade old and lack the semantic and demographic diversity of modern clinical narratives. Large Language Models (LLMs) reach state-of-the-art zero-shot extraction, but their use at enterprise scale is limited by computational cost and by hospital data governance that restricts sending Protected Health Information (PHI) to cloud APIs. We introduce SHIELD (Synthetic Human-annotated Identifier-replaced Entries for Learning and De-identification), a diverse clinical note dataset of 1,381 notes with 10,229 gold-standard PHI spans across 9 categories, built with set-cover diversity sampling across demographic and document-type strata and human-in-the-loop adjudication. We evaluate four LLMs (two proprietary, two open-weight) to establish a performance ceiling on SHIELD, then show that a teacher-student distillation framework transfers these capabilities into locally deployable Small Language Models. Our best distilled model reaches micro-averaged span-level precision of 0.89 and recall of 0.88 while running on standard workstation hardware. It trails its cloud teacher on per-category recall (0.90 vs. 0.81 macro-averaged) but remains competitive given its lower cost and on-premise deployability. Cross-dataset evaluation shows that diversity-trained models generalize well on universal structured PHI categories, while institution-specific entities remain hard to transfer in both directions, which suggests pairing broad-coverage models with specialized models for high-volume, semi-structured note types. We publicly release the SHIELD dataset and the distilled DeBERTa v3 model to provide an accurate, cost-effective de-identification pipeline deployable entirely behind institutional firewalls.

cs.CL↗

arXiv2024

Deep Learning Aided Broadcast Codes with Feedback

Deep learning aided codes have been shown to improve code performance in feedback codes in high noise regimes due to the ability to leverage non-linearity in code design. In the additive white Gaussian broadcast channel (AWGN-BC), the addition of feedback may allow the capacity region to extend far beyond the capacity region of the channel without feedback, enabling higher data rates. On the other hand, there are limited deep-learning aided implementations of broadcast codes. In this work, we extend two classes of deep-learning assisted feedback codes to the AWGN-BC channel; the first being an RNN-based architecture and the second being a lightweight MLP-based architecture. Both codes are trained using a global model, and then they are trained using a more realistic vertical federated learning based framework. We first show that in most cases, using an AWGN-BC code outperforms a linear-based concatenated scheme. Second, we show in some regimes, the lightweight architecture far exceeds the RNN-based code, but in especially unreliable conditions, the RNN-based code dominates. The results show the promise of deep-learning aided broadcast codes in unreliable channels, and future research directions are discussed.

eess.SP↗

arXiv2024

Constant Modulus Waveform Design with Space-Time Sidelobe Reduction for DFRC Systems

Dual-function radar-communication (DFRC) is a key enabler of location-based services for next-generation communication systems. In this paper, we investigate the problem of designing constant modulus multiple-input multiple-output (MIMO) waveforms for DFRC systems. We jointly shape the spatial beam pattern and ambiguity function of the transmit space-time matrix to improve target localization accuracy and enhance target resolution in cluttered environments. For communications, we employ constructive interference (CI)-based precoding, which exploits multi-user and radar-induced interference to enhance MIMO symbol detection. We develop two novel solution algorithms based on majorization-minimization (MM) and the linearized alternating direction method of multipliers (LADMM) principles. For the MM approach, we introduce a novel diagonal majorizer for complex quadratic functions, yielding a tighter surrogate and faster convergence than standard largest eigenvalue-based surrogates. After majorization, we decompose the approximated problem into independent subproblems that can be efficiently solved via parallelizable coordinate descent. To accommodate large MIMO dimensions, we further develop a low-complexity LADMM solution. We combine a biconvex reformulation and first-order proximal approximations to handle the nonconvex quartic objective without requiring costly matrix inversions. We evaluate the performance of the proposed algorithms in comparison to the existing DFRC algorithm. Simulation results demonstrate that the proposed algorithms can substantially enhance target detection and imaging performance due to the reduction of space-time sidelobes.

eess.SP↗

arXiv2023

Fusing Channel and Sensor Measurements for Enhancing Predictive Beamforming in UAV-Assisted Massive MIMO Communications

Cellular-connected unmanned aerial vehicles (UAVs) represent a promising technology for extending the coverage of 5G and 6G networks in a cost-effective manner. Additionally, Massive multiple-input multiple-output (MIMO) serves as an effective solution to interference mitigation in cellular-connected UAV communications. In this letter, we propose a fusion of wireless and sensor data to enhance beam alignment for cellular-connected UAV massive MIMO communications. We develop a predictive beamforming framework, including the frame structure and predictive beamformer. Moreover, we employ an extended Kalman filter (EKF) to integrate channel and sensor data and provide the corresponding state-space and observation models. Simulation results demonstrate that the proposed scheme can improve position/orientation estimation accuracy significantly, leading to higher spectral efficiency.

eess.SP↗

arXiv2023

Robust Non-Linear Feedback Coding via Power-Constrained Deep Learning

The design of codes for feedback-enabled communications has been a long-standing open problem. Recent research on non-linear, deep learning-based coding schemes have demonstrated significant improvements in communication reliability over linear codes, but are still vulnerable to the presence of forward and feedback noise over the channel. In this paper, we develop a new family of non-linear feedback codes that greatly enhance robustness to channel noise. Our autoencoder-based architecture is designed to learn codes based on consecutive blocks of bits, which obtains de-noising advantages over bit-by-bit processing to help overcome the physical separation between the encoder and decoder over a noisy channel. Moreover, we develop a power control layer at the encoder to explicitly incorporate hardware constraints into the learning optimization, and prove that the resulting average power constraint is satisfied asymptotically. Numerical experiments demonstrate that our scheme outperforms state-of-the-art feedback codes by wide margins over practical forward and feedback noise regimes, and provide information-theoretic insights on the behavior of our non-linear codes. Moreover, we observe that, in a long blocklength regime, canonical error correction codes are still preferable to feedback codes when the feedback noise becomes high.

cs.IT↗

arXiv2015

Communicating over Filter-and-Forward Relay Networks with Channel Output Feedback

Relay networks aid in increasing the rate of communication from source to destination. However, the capacity of even a three-terminal relay channel is an open problem. In this work, we propose a new lower bound for the capacity of the three-terminal relay channel with destination-to-source feedback in the presence of correlated noise. Our lower bound improves on the existing bounds in the literature. We then extend our lower bound to general relay network configurations using an arbitrary number of filter-and-forward relay nodes. Such network configurations are common in many multi-hop communication systems where the intermediate nodes can only perform minimal processing due to limited computational power. Simulation results show that significant improvements in the achievable rate can be obtained through our approach. We next derive a coding strategy (optimized using post processed signal-to-noise ratio as a criterion) for the three-terminal relay channel with noisy channel output feedback for two transmissions. This coding scheme can be used in conjunction with open-loop codes for applications like automatic repeat request (ARQ) or hybrid-ARQ.

cs.IT↗

arXiv2007

Hypocomputation

Hypercomputational formal theories will, clearly, be both structurally and foundationally different from the formal theories underpinning computational theories. However, many of the maps that might guide us into this strange realm have been lost. So little work has been done recently in the area of metamathematics, and so many of the previous results have been folded into other theories, that we are in danger of loosing an appreciation of the broader structure of formal theories. As an aid to those looking to develop hypercomputational theories, we will briefly survey the known landmarks both inside and outside the borders of computational theory. We will not focus in this paper on why the structure of formal theory looks the way it does. Instead we will focus on what this structure looks like, moving from hypocomputational, through traditional computational theories, and then beyond to hypercomputational theories.

cs.OH↗