Machine Learning

This paper presents an enhanced methodology for network anomaly detection in Industrial IoT (IIoT) systems using advanced data aggregation and Mutual Information (MI)-based feature selection. The focus is on transforming raw network traffic into meaningful, aggregated forms that capture crucial temporal and statistical patterns. A refined set of 150 features including unique IP counts, TCP acknowledgment patterns, and ICMP sequence ratios was identified using MI to enhance detection accuracy.

This dataset comprises 2 million synthetic samples generated using the Variational Autoencoder-Generative Adversarial Network (VAE-GAN) technique. The dataset is designed to facilitate cardiovascular disease prediction through various demographic, physical, and health-related attributes. It contains essential physiological and behavioral indicators that contribute to cardiovascular health.

Dataset Description The dataset consists of the following features:

LIVE-Viasat Real-World Satellite QoE Database contains 179 videos from real-world streaming, encompassing a range of distortions. Enhanced by a study with 54 participants providing detailed QoE feedback, our work not only provides a rich analysis of the determinants of subjective QoE but also delves into how various streaming impairments influence user behavior, thereby offering a more holistic understanding of user satisfaction.

Large Vision-Language Models (LVLMs) struggle with distractions, particularly in the presence of irrelevant visual or textual inputs. This paper introduces the Irrelevance Robust Visual Question Answering (IR-VQA) benchmark to systematically evaluate and mitigate this ``multimodal distractibility". IR-VQA targets three key paradigms: irrelevant visual contexts in image-independent questions, irrelevant textual contexts in image-dependent questions, and text-only distractions.