Search for: All records

Online Anomaly Detection (OAD) is critical for identifying rare yet important data points in large, dynamic, and complex data streams. A key challenge lies in achieving accurate and consistent detection of anomalies while maintaining computational and memory efficiency. Conventional OAD approaches, which depend on distributional deviations and static thresholds, struggle with model update delays and catastrophic forgetting, leading to missed detections and high false positive rates. To address these limitations, we propose a novel Streaming Anomaly Detection (SAD) method, grounded in a sparse active online learning framework. Our approach uniquely integrates ℓ1,2-norm sparse online learning with CUR decomposition-based active learning, enabling simultaneous fast feature selection and dynamic instance selection. The efficient CUR decomposition further supports real-time residual analysis for anomaly scoring, eliminating the need for manual threshold settings about temporal data distributions. Extensive experiments on diverse streaming datasets demonstrate SAD's superiority, achieving a 14.06% reduction in detection error rates compared to five state-of-the-art competitors. 

High-resolution digital elevation models (HRDEMs), derived from LiDAR, are widely used for mapping hydrographic details in flat terrains. However, artificial flow barriers, particularly from roads, elevate terrain and prematurely end flowlines. Drainage barrier processing (DBP), such as HRDEM excavation, is employed to address this issue. However, there is a gap in quantitatively assessing the impact of DBP on HRDEM-derived flowlines, especially at finer scales. This study fills that gap by quantitatively assessing how DBP improves flowline quality at finer scales. We evaluated HRDEM-derived flowlines that were generated using different flow direction algorithms, developing a framework to measure the effects of flow barrier removal. The results show that the primary factor influencing flowline quality is the presence of flow accumulation artifacts. Quality issues also stem from differences between natural and artificial flow paths, unrealistic flowlines in flat areas, complex canal networks, and ephemeral drainage ways. Notably, the improvement achieved by DBP is demonstrated to be more than 6%, showcasing its efficacy in reducing the impact of flow barriers on hydrologic connectivity. 

Abstract Alternative splicing (AS) of pre-mRNA plays a crucial role in tissue-specific gene regulation, with disease implications due to splicing defects. Predicting and manipulating AS can therefore uncover new regulatory mechanisms and aid in therapeutics design. We introduce TrASPr+BOS, a generative AI model with Bayesian Optimization for predicting and designing RNA for tissue-specific splicing outcomes. TrASPr is a multi-transformer model that can handle different types of AS events and generalize to unseen cellular conditions. It then serves as an oracle, generating labeled data to train a Bayesian Optimization for Splicing (BOS) algorithm to design RNA for condition-specific splicing outcomes. We show TrASPr+BOS outperforms existing methods, enhancing tissue-specific AUPRC by up to 2.4 fold and capturing tissue-specific regulatory elements. We validate hundreds of predicted novel tissue-specific splicing variations and confirm new regulatory elements using dCas13. We envision TrASPr+BOS as a light yet accurate method researchers can probe or adopt for specific tasks. 

Large-scale battery energy storage systems (BESS) play a pivotal role in advancing sustainability through their widespread applications in electrified transportation, power grids, and renewable energy systems. However, achieving optimal power management for these systems poses significant computational challenges. To address this, we propose a scalable approach that partitions the cells of a large-scale BESS into clusters based on state-of-charge (SoC), temperature, and internal resistance. Each cluster is represented by a model that approximates its collective SoC and temperature dynamics and overall power losses during charging and discharging. Using these clusters, we formulate a receding-horizon optimal power control problem to minimize power losses while promoting SoC and temperature balancing. The optimization determines a power quota for each cluster, which is then distributed among its constituent cells. This clustering approach drastically reduces computational costs by working with a smaller number of clusters instead of individual cells, enabling scalability for large-scale BESS. Simulations show a computational overhead reduction of over 60% for small-scale and 98% for large-scale BESS compared to conventional cell-level optimization. Experimental validation using a 20-cell prototype further underscores the approach's effectiveness and practical utility.