Few-shot learning (FSL) is essential in many practical applications. However, the limited training examples make the models more vulnerable to label noise, which can lead to poor generalization capability. To address this critical challenge, we propose a curriculum meta-learning model that employs a novel dual-level class-example sampling strategy to create a robust curriculum for adaptive task distribution formulation and robust model training. The dual-level framework proposes a heuristic class sampling criterion that measures pairwise class boundary complexity to form a class curriculum; it uses effective example sampling through an under-trained proxy model to form an example curriculum. By utilizing both class-level and example-level information, our approach is more robust to handle limited training data and noisy labels that commonly occur in few-shot learning tasks. The model has efficient convergence behavior, which is verified through rigorous convergence analysis. Additionally, we establish a novel error bound through a hierarchical PAC-Bayesian analysis for curriculum meta-learning under noise. We conduct extensive experiments that demonstrate the effectiveness of our framework in outperforming existing noisy few-shot learning methods under various few-shot classification benchmarks. Our code is available at https://github.com/ritmininglab/DCML.

Artificial intelligence (AI) is becoming increasingly prevalent in our society, leading to a growing demand for a skilled workforce in AI. Unfortunately, this demand remains unmet, particularly among smaller institutions and those serving underrepresented groups, due to resource limitations. This initiative offers two hands-on educational activities in Artificial Intelligence and Machine Learning to facilitate the integration of AI/ML concepts into foundational computing and non-computing courses. While primarily aimed at undergraduate students, the materials created can also benefit high school (grades 9–12) and graduate students in various educational settings, including traditional classrooms and outreach or after-school programs. Labs are accessible through a web browser, making them easy to adopt at all institutions, especially those with limited resources. The self-contained and hosted nature of the labs ensures that they can be adopted by institutions facing resource constraints. The complete project material is publicly available on the project website: https://all.rit.edu

Accessibility is a key aspect in ensuring the development of inclusive software. Unfortunately, research demonstrates that a large portion of software today is not created in an accessible manner. Problematically, students may not understand how to create accessible software, additionally misunderstanding the importance of creating accessible software. To fill the gap in accessibility education, we have created a comprehensive collection of labs, collectively referred to as Accessible Learning Labs (ALL). They have the primary objectives of educating participants on how to properly create accessible software, simultaneously illustrating the need to create inclusive and accessible software. In addition, the lab activities enable students to experience the implications of inaccessible software and make repairs based on their experience, making the software in the labs accessible. This will benefit many members of the software engineering community, ranging from beginning-level students to experienced practitioners who want to ensure that they are properly creating accessible and inclusive software. Complete project material is publicly available on the project website: https://all.rit.edu

We propose a novel framework to classify large-scale time series data with long duration. Long time series classification (L-TSC) is a challenging problem because the data often contains a large amount of irrelevant information to the classification target. The irrelevant period degrades the classification performance while the relevance is unknown to the system. This paper proposes an uncertainty-aware multiple instance learning (MIL) framework to identify the most relevant period automatically. The predictive uncertainty enables designing an attention mechanism that forces the MIL model to learn from the possibly discriminant period. Moreover, the predicted uncertainty yields a principled estimator to identify whether a prediction is trustworthy or not. We further incorporate another modality to accommodate unreliable predictions by training a separate model based on its availability and conduct uncertainty aware fusion to produce the final prediction. Systematic evaluation is conducted on the Automatic Identification System (AIS) data, which is collected to identify and track real-world vessels. Empirical results demonstrate that the proposed method can effectively detect the types of vessels based on the trajectory and the uncertainty-aware fusion with other available data modality (Synthetic-Aperture Radar or SAR imagery is used in our experiments) can further improve the detection accuracy.