We study liquid democracy, a collective decision making paradigm that allows voters to transitively delegate their votes, through an algorithmic lens. In our model, there are two alternatives, one correct and one incorrect, and we are interested in the probability that the majority opinion is correct. Our main question is whether there exist delegation mechanisms that are guaranteed to outperform direct voting, in the sense of being always at least as likely, and sometimes more likely, to make a correct decision. Even though we assume that voters can only delegate their votes to better-informed voters, we show that local delegation mechanisms, which only take the local neighborhood of each voter as input (and, arguably, capture the spirit of liquid democracy), cannot provide the foregoing guarantee. By contrast, we design a non-local delegation mechanism that does provably outperform direct voting under mild assumptions about voters.

Label noise in multiclass classification is a major obstacle to the deployment of learning systems. However, unlike the widely used class-conditional noise (CCN) assumption that the noisy label is independent of the input feature given the true label, label noise in real-world datasets can be aleatory and heavily dependent on individual instances. In this work, we investigate the instance-dependent noise (IDN) model and propose an efficient approximation of IDN to capture the instance-specific label corruption. Concretely, noting the fact that most columns of the IDN transition matrix have only limited influence on the class-posterior estimation, we propose a variational approximation that uses a single-scalar confidence parameter. To cope with the situation where the mapping from the instance to its confidence value could vary significantly for two adjacent instances, we suggest using instance embedding that assigns a trainable parameter to each instance. The resulting instance-confidence embedding (ICE) method not only performs well under label noise but also can effectively detect ambiguous or mislabeled instances. We validate its utility on various image and text classification tasks.

Model brittleness is a key concern when deploying deep learning models in real-world medical settings. A model that has high performance at one institution may suffer a significant decline in performance when tested at other institutions. While pooling datasets from multiple institutions and re-training may provide a straightforward solution, it is often infeasible and may compromise patient privacy. An alternative approach is to fine-tune the model on subsequent institutions after training on the original institution. Notably, this approach degrades model performance at the original institution, a phenomenon known as catastrophic forgetting. In this paper, we develop an approach to address catastrophic forgetting based on elastic weight consolidation combined with modulation of batch normalization statistics under two scenarios: first, for expanding the domain from one imaging system's data to another imaging system's, and second, for expanding the domain from a large multi-institutional dataset to another single institution dataset. We show that our approach outperforms several other state-of-the-art approaches and provide theoretical justification for the efficacy of batch normalization modulation. The results of this study are generally applicable to the deployment of any clinical deep learning model which requires domain expansion.

This article describes research on claim verification carried out using a multiple GAN-based model. The proposed model consists of three pairs of generators and discriminators. The generator and discriminator pairs are responsible for generating synthetic data for supported and refuted claims and claim labels. A theoretical discussion about the proposed model is provided to validate the equilibrium state of the model. The proposed model is applied to the FEVER dataset, and a pre-trained language model is used for the input text data. The synthetically generated data helps to gain information which helps the model to perform better than state of the art models and other standard classifiers.

Self-driving company TuSimple Inc. unveiled paperwork for its initial public offering Tuesday showing it has lost more than $300 million over the past three years in the race to be the first to launch fully autonomous long-haul trucks. TuSimple had already filed confidentially for an IPO, The Wall Street Journal reported, and the Tuesday filing offered the public the first detailed look at a startup that has attracted more funding than many of its Silicon Valley counterparts and maintained split operations in California and China. Its China connections have caught the attention of U.S. regulators. The Committee on Foreign Investment in the U.S., or Cfius, has identified TuSimple as a company meriting review because of its ties to China and because autonomous driving technology is considered a critical technology for the Department of Defense. Cfius alerted TuSimple this month that it was probing a Chinese investment in the company from 2017, according to the IPO filing.

It is profoundly difficult to grapple with risks whose stakes may include the global collapse of civilisation, or even the extinction of humanity. The pandemic has shattered our illusions of safety and reminded us that despite all the progress made in science and technology, we remain vulnerable to catastrophes that can overturn our entire way of life. These are live possibilities, not mere hypotheses, and our governments will have to confront them. As Britain emerges from Covid-19, it could find itself at the forefront of the response to future disasters. The government's recent integrated review, Britain's taking of the G7 presidency and the Cop26 climate conference, which will be hosted in Glasgow later this year, are all occasions to address global crises. But in order to ensure that the UK really is prepared, we need to first identify the biggest risks that we face in the coming decades.

Arabic poses a set of unique speech recognition challenges due to its rich dialectal variety.

The world is facing enormous challenges, ranging from climate change to extreme poverty. The 2030 Agenda for Sustainable Development and its 17 Sustainable Development Goals (SDGs)a were adopted by United Nations Member States in 2015 as an operational framework to address these challenges. The SDGs include No Poverty, Quality Education, Gender Equality, Peace, Justice and Strong Institutions, among others, as well as a meta goal on Partnerships for the Goals. Despite limitations,7 the SDGs form a rare global consensus of all 193 UN member states on where we should collectively be heading. Goals are meaningless without a way to track their progress. Data on the SDGs and the associated indicatorsb are often outdated or unavailable, hindering progress during the Decade of Action leading up to 2030.c

Dana Abu Ali is a software engineer at MEA Clients Center at IBM Dubai.

The Qatar Computing Research Institute (QCRI) is one of three national research institutes established in 2010 by Qatar Foundation (QF) for education, science and community development. It operates under the umbrella of Hamad Bin Khalifa University and is steered operationally by the Research, Development, and Innovation (RDI) division, which was established within QF to oversee the three national research institutes' day-to-day operations. In this capacity, RDI provides high-level planning, coordination, and oversight to further the institutes' research priorities. QCRI was created with a mandate to support Qatar's transformation from a carbon economy to a knowledge-based economy. In doing so, it fulfills Qatar Foundation's overarching objectives of enabling national and regional change.