FSU Computer Science Assistant Professor, Dr. Shayok Chakraborty, has received the prestigious AWS Machine Learning Research Award for his project “Active Learning with Imperfect Oracles”. Active learning algorithms automatically identify the salient and exemplar samples from large amounts of unlabeled data and tremendously reduce human annotation effort in inducing a machine learning model. In a traditional active learning setup, the labeling oracles are assumed to be infallible, that is, they always provide correct answers (in terms of class labels) to the queried unlabeled instances. However, in real-world applications, oracles are often imperfect and provide incorrect label annotations. For instance, in a crowd-sourcing platform there are a variety of workers with different experience and expertise, providing annotations of varied quality. The goal of this research is to develop novel active learning algorithms under such real-world constraints. Such a technology can be immensely useful in services like the Amazon SageMaker Ground Truth, which uses active learning to create training datasets for inducing predictive models.
The award includes cash funding in the amount of $35,000 and AWS Promotional Credits in the amount of $25,000 for one year.