The field of sequence alignment and shape characterization is witnessing significant developments, driven by the need for more accurate and efficient methods. Researchers are exploring new approaches to improve the alignment of sequences, including the use of optimal algorithms and the characterization of rotational symmetries of shapes embedded in curved surfaces. Furthermore, there is a growing interest in developing evaluation frameworks that prioritize clinical priorities, such as calibration and robustness to distributional shifts, in machine learning-based decision support systems. Another area of focus is the learning of global ground metrics for comparing probability distributions, which has applications in scRNA data analysis. Notable papers in this area include:
- A study that introduces surface Minkowski tensors to characterize rotational symmetries of shapes embedded in curved surfaces, providing a novel approach for shape analysis.
- A paper that proposes a principled evaluation framework for selecting calibrated thresholded classifiers that explicitly accounts for uncertainty in class prevalences and domain-specific cost asymmetries.
- A work that presents a novel approach for learning global ground metrics for arbitrary distributions over a shared metric space, enabling more accurate optimal transport distances and improved performance in embedding, clustering, and classification tasks.