Research Challenge

Current research addresses significant challenges in various AI subfields, focusing on improving the capabilities and trustworthiness of large language models (LLMs), enhancing anomaly detection in time series data, and bolstering the robustness and explainability of deep neural networks (DNNs). Key areas of investigation include refining LLM fine-tuning techniques, developing more effective model-based anomaly detection methods, and exploring neurosymbolic AI approaches for improved reasoning. These advancements are crucial for expanding the reliable application of AI across diverse domains, from scientific discovery and industrial processes to healthcare and autonomous systems.