Applications of machine learning in healthcare often require working with time-to-event prediction tasks including prognostication of an adverse event, re-hospitalization or death. Such outcomes are typically subject to censoring due to loss of follow up. Standard machine learning methods cannot be applied in a straightforward manner to datasets with censored outcomes. In this paper, we present auton-survival, an open-source repository of tools to streamline working with censored time-to-event or survival data. auton-survival includes tools for survival regression, adjustment in the presence of domain shift, counterfactual estimation, phenotyping for risk stratification, evaluation, as well as estimation of treatment effects. Through real world case studies employing a large subset of the SEER oncology incidence data, we demonstrate the ability of auton-survival to rapidly support data scientists in answering complex health and epidemiological questions.

Currently, deep neural networks (DNNs) are widely adopted in different applications. Despite its commercial values, training a well-performed DNN is resource-consuming. Accordingly, the well-trained model is valuable intellectual property for its owner. However, recent studies revealed the threats of model stealing, where the adversaries can obtain a function-similar copy of the victim model, even when they can only query the model. In this paper, we propose an effective and harmless model ownership verification (MOVE) to defend against different types of model stealing simultaneously, without introducing new security risks. In general, we conduct the ownership verification by verifying whether a suspicious model contains the knowledge of defender-specified external features. Specifically, we embed the external features by tempering a few training samples with style transfer. We then train a meta-classifier to determine whether a model is stolen from the victim. This approach is inspired by the understanding that the stolen models should contain the knowledge of features learned by the victim model. In particular, we develop our MOVE method under both white-box and black-box settings to provide comprehensive model protection. Extensive experiments on benchmark datasets verify the effectiveness of our method and its resistance to potential adaptive attacks. The codes for reproducing the main experiments of our method are available at \url{https://github.com/THUYimingLi/MOVE}.

It has been around in our cultures one form or another since the times of the Ancient Greeks and their myths, through to Frankenstein, and Asimov. This long and storied history cannot take away from the fact that AI is now front and center in our world. AI technology is both for Ericsson and our customers a key business enabler. Looking back at the history of AI, we see a recurring theme. Using AI wrongly, or without due diligence can lead to a widespread escalation of problems on many fronts.

Kabul, Afghanistan – The news of the killing of al-Qaeda chief Ayman al-Zawahiri slowly made its way through the Afghan capital. For many Afghans, it came as a complete surprise. The announcement by the United States of a "precision" drone attack that killed the elusive 71-year-old al-Qaeda leader came in Kabul in the early hours of Tuesday. As the day advanced, more details started to trickle in. However, in a sign of the growing fears over the freedom of speech under a Taliban government, many city residents seemed hesitant to talk about the killing of al-Zawahiri, who had a reward of $25m on his head for the 9/11 attacks.

Role requiring'No experience data provided' months of experience in None We're looking for a Build and Release Engineer to join our team as a master of packages and containers who always delivers the goods. A key member of the CI/CD pipeline from planning all the way through deployment, you'll collaborate with other skilled engineers to identify technical needs, develop solutions, and deploy them using the latest tools available. Sure, you'll use your head to build and maintain the tools, infrastructure, and processes that directly impact our development teams and customers. But you're a developer at heart, and this is a role in which you'll craft code aplenty and keep your finger on the pulse of modern software build engineering practices. Inductive Automation is an innovation company. We are champions for industrial automation software, and we believe in building sensible solutions that provide value for our customers. Our workforce shares a passion for technology.

The ideal candidate will be responsible for architecting automation solutions for QA services, along with providing support for implementation. Candidate will be expected to effectively lead, monitor and improve automation service creation and business growth.

At GoDaddy the future of work looks different for each team. Some teams work in the office full-events or offsites. Your hiring manager can share more about this role's hybrid or remote time, others have a hybrid arrangement (they work remotely some days and in the office some days) and some work entirely remotely. Remote: This is a remote position, so you'll be working remotely from your home. You may occasionally visit a GoDaddy office to meet with your team for events or offsites.

New York police must now comply with a public records request related to its use of facial recognition and other surveillance on protestors. A judge has ordered the New York Police Department to release documents pertaining to its monitoring of Black Lives Matters protests during the summer of 2020, requiring it to release 2,700 emails and other documents to the public or state why it fall"and/or allege with specificity that each document falls within one of the enumerated exemptions of Public Officers Law." The NYPD previously rejected a Freedom of Information Law (FOIL) request by Amnesty International and the Surveillance Technology Oversight Project for records related to its use of facial recognition and surveillance tools on activists (as well as a subsequent appeal to that FOIL request), leading both groups to sue the law enforcement organization last year. The police agency has argued that the records request would cover over 30 million documents, and that following through would be "unreasonably burdensome." In a ruling issued on Friday, New York Supreme Court Justice Lawrence Love rejected the NYPD's reasoning.

With recent improvements in natural language generation (NLG) models for various applications, it has become imperative to have the means to identify and evaluate whether NLG output is only sharing verifiable information about the external world. In this work, we present a new evaluation framework entitled Attributable to Identified Sources (AIS) for assessing the output of natural language generation models, when such output pertains to the external world. We first define AIS and introduce a two-stage annotation pipeline for allowing annotators to appropriately evaluate model output according to AIS guidelines. We empirically validate this approach on generation datasets spanning three tasks (two conversational QA datasets, a summarization dataset, and a table-to-text dataset) via human evaluation studies that suggest that AIS could serve as a common framework for measuring whether model-generated statements are supported by underlying sources. We release guidelines for the human evaluation studies.

Concerns about the environmental impacts of advanced technologies such as AI is prompting debate over whether computing-intensive applications and the chips that power them need to be regulated. That's according to experts speaking during the "Advancing Technology for a Sustainable Planet" conference hosted by the Stanford Institute for Human-Centered Artificial Intelligence and the Stanford Woods Institute for the Environment. At the same time, many of these technologies also help companies meet sustainability goals -- meaning companies need balance between quickly adopting and scaling emerging technologies and understanding how those technologies could affect the company's overall environmental impact. Environmental impact depends on scale -- particularly for a technology like artificial intelligence, said Peter Henderson, a Ph.D student in computer science at Stanford University, during a conference panel session. Companies often optimize AI algorithms to address energy use and carbon emission concerns before deploying a machine learning model.