Abstract Time series data is pervasive across various domains, from finance and economics to healthcare and environmental science. The application of Machine Learning (ML) techniques to time series data has opened new avenues for enhancing predictive accuracy and uncovering hidden trends that traditional methods might overlook. In this talk, we will start with two machine learning applications in time series data, one on gait analysis using accelerometer data and one on preclinical Alzheimer’s disease detection using EEG signals from word repetition tests. Beyond these applications, the talk will address the growing field of time series explainability, including its unique challenges and our recent advances.