To build an Internet of Things (IoT) infrastructure that provides flood susceptibility forecasts for granular geographic levels, an extensive network of IoT weather sensors in local regions is crucial. However, these IoT devices may exhibit anomalistic behavior due to factors such as diminished signal strength, physical disturbance, low battery life, and more. To ensure that incorrect readings are identified and addressed appropriately, we devise a novel method for multi-stream sensor data verification and anomaly detection. Our method uses time-series anomaly detection to identify incorrect readings. We expand on the state-of-the-art by incorporating sensor fusion mechanisms between nearby devices to improve anomaly detection ability. Our system pairs nearby devices and fuses them by creating a new time series with the difference between the corresponding readings. This new time series is then input into a time-series anomaly detection model which identifies if any readings are anomalistic. By testing our system with nine different machine learning anomaly detection methods on synthetic data based on one year of real weather data, we find that our system outperforms the previous anomaly detection methods by improving F1-Score by 10.8%.
The motion of leukocytes is significant in studying the inflammation response of the immune system. In inflammation conditions, some leukocytes slow down and eventually adhere to vessel walls. With many cells moving at a variety of speeds, collisions occur. These collisions result in abrupt changes in the motion and appearance of leukocytes. In this paper, we propose a novel method of tracking multiple cells undergoing collision by modeling the collision states of cells and testing multiple hypotheses of their motion and appearance. We demonstrate the results on in vivo intravital microscopy image sequences. The proposed method results in 20% more accurate cell position and 82% longer tracks than a previous approach.
As computer science is integrated into a wider variety of fields, block-based programming languages like Snap!, which assemble code with visual blocks rather than text syntax, are increasingly used to teach computational thinking (CT) to students from diverse backgrounds. Although automated evaluators (autograders) for programming assignments usually focus on runtime efficiency and output accuracy, effective evaluation of a student's CT skills requires assessing coding best practices, such as decomposition, abstraction, and algorithm design. While autograders are commonplace for text languages like Python, we present a machine learning approach to assess how effectively block-based code demonstrates understanding of CT fundamentals. Our dataset consists of Snap! programs written by students new to coding and evaluated by instructors using a CT rubric. We explore how to best transform these programs into low-dimensional features to allow encapsulation and repetition patterns to emerge. Experimentation involves comparing the effectiveness of a suite of clustering models and similarity metrics by analyzing how directly automated feedback correlates to the course staff's manual evaluation. Lastly, we demonstrate the practical application of the autograder in a classroom setting and discuss scalability and feasibility in other domains of CS education.