On December 16, 2025, we held our Ideal R&D Seminar with Assoc. Prof. Dr. Yekta Said Can. Dr. Can presented a comprehensive approach to how Large Language Models (LLMs) can be effectively used in two different data fields in his talk titled “LLM-Supported Time Series Analysis and Forecasting: From Physiological Signals to Financial Price Forecasting.”

The first part of the two-part seminar examined an explainable stress recognition infrastructure based on PPG data obtained from wearable devices used in daily life. Dr. Can provided information on how HRV-based manually extracted features are interpreted by physiological threshold rules generated by LLM. In the second part, the LLM-supported reasoning approach was extended to time series forecasting and price prediction; ARIMA, Prophet, LSTM, and Transformer architectures were compared using a real stock market prediction example. He applied a holistic perspective that combines signal processing, model explainability, and prediction-focused AI design.

Assistant Professor Yekta Said Can is a Lecturer/Assistant Professor at Augsburg University, focusing on deep learning, wearable technologies, affective computing, and physiological signal processing. His work has been published in leading journals such as IEEE Transactions on Affective Computing, IEEE Access, ACM ICIE, and Frontiers in Psychology, making significant contributions to multi-modal self-regulated learning, explainable stress recognition, and wearable emotion tracking.