Congratulations to Dr Li Xiaoli, Investigator from A*STAR’s Centre for Frontier AI Research (CFAR) for receiving the Institute of Electrical and Electronics Engineers (IEEE) fellow for his contributions to machine learning models.

The IEEE Fellow is one of the most prestigious honours of the IEEE and is a testament to the individual’s significant impact on engineering, science, and technology in our society.

Dr Li’s research interests include artificial intelligence (AI), data mining, machine learning, and bioinformatics. As the Principal Investigator of the AI Singapore (AISG) grant on continuous learning, he has driven several collaborations with aerospace, semiconductor, and manufacturing companies to generate domain-specific solutions and real-world impacts.

Dr Li’s most noteworthy accomplishment involves creating advanced machine learning models to process time-series sensor data. He pioneered a novel representation learning method to automatically extract features from high dimensional sensor signals and apply it to various practical applications, such as sensor-based activity recognition, EEG-based sleep stage classification, equipment health diagnosis and prognosis, and the prediction of remaining useful life. To tackle key challenges in time-series data analysis, Dr Li proposed innovative deep learning models in the fields of representation learning, adversarial domain adaptation, self-supervised learning, knowledge distillation, and structure preserving oversampling. His groundbreaking work in sensor data learning has received more than 3,000 citations and has been recognised with two best paper awards at prestigious IEEE international conferences (ICPHM, ICIEA) and one best performance award at the EU opportunity activity recognition challenge.

In addition, Dr Li’s thorough and rigorous approach to developing models for Positive Unlabeled Learning (PU learning) has made him one of the first researchers to formulate a PU learning problem, with the term PU learning coined in his paper. Widely regarded as the foundation of the field, his seminal papers on PU learning have received over 3,000 citations. Besides developing algorithms featured in four most influential PU learning papers, Dr Li also constructed the PU learning tool on his own and made it accessible to the scientific community, garnering more than 3,000 downloads to date.

Read more about IEEE Fellow Program.