Search for: All records

Stress affects physical and mental health, and wearable devices have been widely used to detect daily stress through physiological signals. However, these signals vary due to factors such as individual differences and health conditions, making generalizing machine learning models difficult. To address these challenges, we present Human Heterogeneity Invariant Stress Sensing (HHISS), a domain generalization approach designed to find consistent patterns in stress signals by removing person-specific differences. This helps the model perform more accurately across new people, environments, and stress types not seen during training. Its novelty lies in proposing a novel technique called person-wise sub-network pruning intersection to focus on shared features across individuals, alongside preventing overfitting by leveraging continuous labels while training. The present study focuses on people with opioid use disorder (OUD)---a group where stress responses can change dramatically depending on the presents of opioids in their system, including daily timed medication for OUD (MOUD). Since stress often triggers cravings, a model that can adapt well to these changes could support better OUD rehabilitation and recovery. We tested HHISS on seven different stress datasets---four which we collected ourselves and three public datasets. Four are from lab setups, one from a controlled real-world driving setting, and two are from real-world in-the-wild field datasets with no constraints. The present study is the first known to evaluate how well a stress detection model works across such a wide range of data. Results show HHISS consistently outperformed state-of-the-art baseline methods, proving both effective and practical for real-world use. Ablation studies, empirical justifications, and runtime evaluations confirm HHISS's feasibility and scalability for mobile stress sensing in sensitive real-world applications. 

The advancement in deep learning and internet-of-things have led to diverse human sensing applications. However, distinct patterns in human sensing, influenced by various factors or contexts, challenge the generic neural network model's performance due to natural distribution shifts. To address this, personalization tailors models to individual users. Yet most personalization studies overlook intra-user heterogeneity across contexts in sensory data, limiting intra-user generalizability. This limitation is especially critical in clinical applications, where limited data availability hampers both generalizability and personalization. Notably, intra-user sensing attributes are expected to change due to external factors such as treatment progression, further complicating the challenges. To address the intra-user generalization challenge, this work introduces CRoP, a novel static personalization approach. CRoP leverages off-the-shelf pre-trained models as generic starting points and captures user-specific traits through adaptive pruning on a minimal sub-network while allowing generic knowledge to be incorporated in remaining parameters. CRoP demonstrates superior personalization effectiveness and intra-user robustness across four human-sensing datasets, including two from real-world health domains, underscoring its practical and social impact. Additionally, to support CRoP's generalization ability and design choices, we provide empirical justification through gradient inner product analysis, ablation studies, and comparisons against state-of-the-art baselines. 

Polyploidy and subsequent post-polyploid diploidization (PPD) are key drivers of plant genome evolution, yet their contributions to evolutionary success remain debated. Here, we analyze the Malvaceae family as an exemplary system for elucidating the evolutionary role of polyploidy and PPD in angiosperms, leveraging 11 high-quality chromosome-scale genomes from all nine subfamilies, including newly sequenced, near telomere-to-telomere assemblies from four of these subfamilies. Our findings reveal a complex reticulate paleoallopolyploidy history early in the diversification of the Malvadendrina clade, characterized by multiple rounds of species radiation punctuated by ancient allotetraploidization (Mal-β) and allodecaploidization (Mal-α) events around the Cretaceous–Paleogene (K–Pg) boundary. We further reconstruct the evolutionary dynamics of PPD and find a strong correlation between dysploidy rate and taxonomic richness of the paleopolyploid subfamilies (R^2 ≥ 0.90, P < 1e-4), supporting the “polyploidy for survival and PPD for success” hypothesis. Overall, our study provides a comprehensive reconstruction of the evolutionary history of the Malvaceae and underscores the crucial role of polyploidy–dysploidy waves in shaping plant biodiversity. 

Conventional directed evolution methods offer the ability to select bioreceptors with high binding affinity for a specific target in terms of thermodynamic properties. However, there is a lack of analogous approaches for kinetic selection, which could yield affinity reagents that exhibit slow off-rates and thus remain tightly bound to targets for extended periods. Here, we describe an in vitro directed evolution methodology that uses the nuclease flap endonuclease 1 to achieve the efficient discovery of aptamers that have slow dissociation rates. Our nuclease-assisted selection strategy can yield specific aptamers for both small molecules and proteins with off-rates that are an order of magnitude slower relative to those obtained with conventional selection methods while still retaining excellent overall target affinity in terms of thermodynamics. This new methodology provides a generalizable approach for generating slow off-rate aptamers for diverse targets, which could, in turn, prove valuable for applications including molecular devices, bioimaging, and therapy. 

AI's widespread integration has led to neural networks (NN) deployment on edge and similar limited-resource platforms for safety-critical scenarios. Yet, NN's fragility raises concerns about reliable inference. Moreover, constrained platforms demand compact networks. This study introduces VeriCompress, a tool that automates the search and training of compressed models with robustness guarantees. These models are well-suited for safety-critical applications and adhere to predefined architecture and size limitations, making them deployable on resource-restricted platforms. The method trains models 2-3 times faster than the state-of-the-art approaches, surpassing them by average accuracy and robustness gains of 15.1 and 9.8 percentage points, respectively. When deployed on a resource-restricted generic platform, these models require 5-8 times less memory and 2-4 times less inference time than models used in verified robustness literature. Our comprehensive evaluation across various model architectures and datasets, including MNIST, CIFAR, SVHN, and a relevant pedestrian detection dataset, showcases VeriCompress's capacity to identify compressed verified robust models with reduced computation overhead compared to current standards. This underscores its potential as a valuable tool for end users, such as developers of safety-critical applications on edge or Internet of Things platforms, empowering them to create suitable models for safety-critical, resource-constrained platforms in their respective domains.