News articles constitute a valuable resource for opinion mining, as they contain important perspectives related to the subject matter they cover. In this paper, we explore how aspect-based sentiment analysis might help in understanding the public discourse surrounding agricultural biotechnologies in Africa. We introduce BioMAISx, the first English language dataset composed of direct quotes pertaining to agricultural biotechnologies extracted from a curated list of Africa-based news sources. We have identified and labelled entities related to key aspects of agricultural biotechnologies, providing valuable insights into public discourse. This dataset can aid in identifying challenges, improving public discourse, and monitoring the perception of agricultural biotechnologies, thus contributing to informed decision-making.

Laser scanning microscopy is a powerful imaging modality ideal for monitoring spatial and temporal dynamics in both in vitro and in vivo models. To accurately resolve dynamic changes, particular to the neuroimaging field, fast acquisition rates are in great need. Unfortunately, the video-rate acquisition required to capture these changes comes with a trade-off between resolution, high spatial distortion, and low signal-to-noise ratio due to the electronics and Poisson noise. By combining microscopy fast acquisition methods with a Generative Adversarial Network (GAN), we show here, for the first time, that video-rate image acquisition, up to 20x the speed of equivalent standard high resolution acquisition, can be obtained without significant reduction in image quality. Specifically, we present a GAN based training approach that is able to simultaneously 1) super-resolve, 2) denoise and 3) correct distortion on fast scanning acquisition microscopy images. In addition, we show that our method generalizes on unseen data, requires minimal ground truth images for training and can easily be fine-tuned on different biological samples.

Atrial Fibrillation is a heart condition characterized by erratic heart rhythms caused by chaotic propagation of electrical impulses in the atria, leading to numerous health complications. State-of-the-art models employ complex algorithms that extract expert informed features to improve diagnosis. In this note, we demonstrate how topological features can be used to help accurately classify single lead electrocardiograms. Via delay embeddings, we map electrocardiograms onto high-dimensional point-clouds that convert periodic signals to algebraically computable topological signatures. We derive features from persistent signatures, input them to a simple machine learning algorithm, and benchmark its performance against winning entries in the 2017 Physionet Computing in Cardiology Challenge.

The purpose of this panel is to discuss the creation and implementation of a data science degree program at the undergraduate level. The panel includes representatives from three different universities that each offers an undergraduate degree in Data Science as of fall 2013. We plan to share information on the logistics of how the data science programs came to exist at each of our schools as well as encourage a robust interactive discussion about the future of data science education at the undergraduate level.

Ideas from the image processing literature have recently motivated a new set of clustering algorithms that rely on the concept of total variation. While these algorithms perform well for bi-partitioning tasks, their recursive extensions yield unimpressive results for multiclass clustering tasks. This paper presents a general framework for multiclass total variation clustering that does not rely on recursion. The results greatly outperform previous total variation algorithms and compare well with state-of-the-art NMF approaches.

Unsupervised clustering of scattered, noisy and high-dimensional data points is an important and difficult problem. Continuous relaxations of balanced cut problems yield excellent clustering results. This paper provides rigorous convergence results for two algorithms that solve the relaxed Cheeger Cut minimization. The first algorithm is a new steepest descent algorithm and the second one is a slight modification of the Inverse Power Method algorithm \cite{pro:HeinBuhler10OneSpec}. While the steepest descent algorithm has better theoretical convergence properties, in practice both algorithm perform equally. We also completely characterize the local minima of the relaxed problem in terms of the original balanced cut problem, and relate this characterization to the convergence of the algorithms.