Kangshuo Li's repositories
GR5398_IoT
The project is aimed at developing new tools for classifying videos of human-machine interactions in the Internet-of-Things (IOT) domain. Namely, given videos of humans interacting with IoT devices (e.e., smart appliances such as fridge, toaster, washing machines, Alexa, etc), the aim is to (1) design predictive video features, which (2) can be extracted efficiently in real-time to classify videos in terms of the activity being performed (opening or closing a fridge, loading or unloading a washing machine, etc.). The grand goal and motivation for the work is to generate labels for IoT network traffic, simply by training cameras onto IoT devices in the various IoT labs across US universities. Thus, the project aims to solve a main bottleneck in research at the intersection of Machine Learning and IoT, namely, the scarcity of labeled IoT traffic data to solve ML problems such as activity and anomaly detection using supervised or unsupervised detection procedures.
chia
CHIA: Concept Hierarchies for Incremental and Active Learning
ML_IS_Prediction
Machine Learning-based Prediction of Infarct Size in Patients with ST-segment Elevation Myocardial Infarction: A Multi-center Study