Track: Emerging Technologies and RTC |
| Robust Machine Learning for Real-world Time-series Applications |
| Wearables, smartphones, and continuous sensing applications are ubiquitous today. While they have achieved tremendous success thanks to improved computation, smaller form factors, and advanced sensing capabilities, the analytics for these applications still lag behind those in the language and vision domains. This disparity arises because real-world time-series data are imperfect—typically noisy, frequently encounter missing values, and are generally non-interpretable in their raw format, among other challenges. In this talk, I will present two practical time-series frameworks: (1) phase-based generalizable representation learning for nonstationary time-series classification and (2) learning from heterogeneous and abruptly missing sensors through a robust and efficient multimodal learning approach. I will also briefly illustrate some human-centric applications such as fatigue monitoring and speech disfluency detection that face challenges related to data, labels, and resource limitations, along with some techniques to address these issues through machine learning. |
|
| Presentation Video |
| Presentation Notes |
|
Mohapatra-RobustML.pdf |
Real Time Communications Conference & Expo at Illinois Tech
IEEE International Conference
Follow Us
Share This
