The rapidly emerging field of deep learning-based computational pathology has demonstrated promise in developing objective prognostic models from histology whole slide images. However, most prognostic models are either based on histology or genomics alone and do not address how histology and genomics can be integrated to develop joint image-omic prognostic models. Additionally identifying explainable morphological and molecular descriptors from these models that govern such prognosis is of interest. We used multimodal deep learning to integrate gigapixel whole slide pathology images, RNA-seq abundance, copy number variation, and mutation data from 5,720 patients across 14 major cancer types. Our interpretable, weakly-supervised, multimodal deep learning algorithm is able to fuse these heterogeneous modalities for predicting outcomes and discover prognostic features from these modalities that corroborate with poor and favorable outcomes via multimodal interpretability. We compared our model with unimodal deep learning models trained on histology slides and molecular profiles alone, and demonstrate performance increase in risk stratification on 9 out of 14 cancers. In addition, we analyze morphologic and molecular markers responsible for prognostic predictions across all cancer types. All analyzed data, including morphological and molecular correlates of patient prognosis across the 14 cancer types at a disease and patient level are presented in an interactive open-access database (http://pancancer.mahmoodlab.org) to allow for further exploration and prognostic biomarker discovery. To validate that these model explanations are prognostic, we further analyzed high attention morphological regions in WSIs, which indicates that tumor-infiltrating lymphocyte presence corroborates with favorable cancer prognosis on 9 out of 14 cancer types studied.

To enable the reusability of massive scientific datasets by humans and machines, researchers aim to create scientific datasets that adhere to the principles of findability, accessibility, interoperability, and reusability (FAIR) for data and artificial intelligence (AI) models. This article provides a domain-agnostic, step-by-step assessment guide to evaluate whether or not a given dataset meets each FAIR principle. We then demonstrate how to use this guide to evaluate the FAIRness of an open simulated dataset produced by the CMS Collaboration at the CERN Large Hadron Collider. This dataset consists of Higgs boson decays and quark and gluon background, and is available through the CERN Open Data Portal. We also use other available tools to assess the FAIRness of this dataset, and incorporate feedback from members of the FAIR community to validate our results. This article is accompanied by a Jupyter notebook to facilitate an understanding and exploration of the dataset, including visualization of its elements. This study marks the first in a planned series of articles that will guide scientists in the creation and quantification of FAIRness in high energy particle physics datasets and AI models.

Humanitarian mapping from space with machine learning helps policy-makers to timely and accurately identify people in need. However, recent concerns around fairness and transparency of algorithmic decision-making are a significant obstacle for applying these methods in practice. In this paper, we study if humanitarian mapping approaches from space are prone to bias in their predictions. We map village-level poverty and electricity rates in India based on nighttime lights (NTLs) with linear regression and random forest and analyze if the predictions systematically show prejudice against scheduled caste or tribe communities. To achieve this, we design a causal approach to measure counterfactual fairness based on propensity score matching. This allows to compare villages within a community of interest to synthetic counterfactuals. Our findings indicate that poverty is systematically overestimated and electricity systematically underestimated for scheduled tribes in comparison to a synthetic counterfactual group of villages. The effects have the opposite direction for scheduled castes where poverty is underestimated and electrification overestimated. These results are a warning sign for a variety of applications in humanitarian mapping where fairness issues would compromise policy goals.

We consider a model of committee elections, where the goal is to select a fixed-size subset of objects based on the preferences of a group of individuals. The objects and the individuals are typically referred to as the candidates and the voters, respectively, and we follow this convention in our paper. However, the candidates do not need to represent humans. For example, this model describes (1) the problem of locating public facilities--there the candidates correspond to possible physical locations where the facilities can be built [Farahani and Hekmatfar, 2009, Skowron et al., 2016], (2) the problem of presenting results by a search engine in response to a user query--there, the candidates are web-pages, and voters are potential users searching for a given query [Skowron et al., 2017], (3) the problem of selecting validators in the blockchain, where the candidates are the users of the protocol [Cevallos and Stewart, 2020, Burdges et al., 2020]. For more examples that fall into the category of committee elections we refer to the recent book chapter [Faliszewski et al., 2017] and to the recent survey [Lackner and Skowron, 2020]. In numerous applications that fit the model of committee elections it is critical to select a subset of candidates, hereinafter called a committee, in a fair and proportional manner. Proportionality is one of fundamental requirements of methods for selecting representative bodies, such as parliaments, faculty boards, etc.

This work licensed under a Creative Commons Attribution 4.0 license. That means you can use it any way you like, including commercially, provided that you attribute it to me, Cory Doctorow, and include a link to pluralistic.net.

The National Institute of Standards and Technology is seeking public input on what to include in forthcoming guidance that will set rules of the road for fielding trustworthy artificial intelligence in and out of government. NIST, following the recommendations of the National Security Commission on AI, is working on an AI Risk Management Framework that will set voluntary standards for agencies and industries to consider when adopting AI solutions. NIST, in a request for information posted Wednesday, said the upcoming framework will define trustworthy AI in terms of transparency, fairness and accountability. The agency plans to release the framework as a "living document" that adapts to changes in technology and practices. "Defining trustworthiness in meaningful, actionable, and testable ways remains a work in progress," the agency wrote in its RFI.

We are a publication devoted to covering the business and enterprise applications of robotic process automation (RPA), artificial intelligence (AI), machine learning, cognitive computing, and other aspects of automation.

A small company developing an implantable brain computer interface to help treat conditions like paralysis has received the go-ahead from the Food and Drug Administration (FDA) to kick off clinical trials of its flagship device later this year. New York-based Synchron announced Wednesday it has received FDA approval to begin an early feasibility study of its Stentrode implant later this year at Mount Sinai Hospital with six human subjects. The study will examine the safety and efficacy of its motor neuroprosthesis in patients with severe paralysis, with the hopes the device will allow them to use brain data to "control digital devices and achieve improvements in functional independence." "Patients begin using the device at home soon after implantation and may wirelessly control external devices by thinking about moving their limbs. The system is designed to facilitate better communication and functional independence for patients by enabling daily tasks like texting, emailing, online commerce and accessing telemedicine," the company said in a release.

Each of the 11 new research institutes, which are headed by major research universities, will ... [ ] receive about $20 million over five years. The National Science Foundation (NSF) announced today that it was funding 11 new National Artificial Intelligence (AI) Research Institutes. The latest grants follow a first round of seven AI research institutes established in 2020. The new investment totals $220 million and expands the network of these institutes to a total of 40 states and the District of Columbia, according to NSF. "I am delighted to announce the establishment of new NSF National AI Research Institutes as we look to expand into all 50 states," said National Science Foundation Director Sethuraman Panchanathan in the agency's news release. "These institutes are hubs for academia, industry and government to accelerate discovery and innovation in AI. Inspiring talent and ideas everywhere in this important area will lead to new capabilities that improve our lives from medicine to entertainment to transportation and cybersecurity and position us in the vanguard of competitiveness and prosperity."

The US military is conducting tests involving artificial intelligence, cloud computing, and sensors, that could give it the ability to predict events "days in advance". The system, called Global Information Dominance Experiments (GIDE), has been tested in three times according to General Glen VanHerck, Commander of United States Northern Command and North American Aerospace Defense Command. These two organisations are part of eleven unified combatant commands of the Department of Defense, which also include Space Command, Cyber Command, and various commands for geographic areas. "The threats we face and the pace of change in the geostrategic environment continues to advance at really alarming rates. We've entered a era of new and renewed strategic competition, and this time, we're facing two peer competitors, both nuclear-armed, that are competing against us on a daily basis", General VanHerck said during a press briefing.