Building machine learning models that are fair with respect to an unprivileged group is a topical problem. Modern fairness-aware algorithms often ignore causal effects and enforce fairness through modifications applicable to only a subset of machine learning models. In this work, we propose a new definition of fairness that incorporates causality through the Controlled Direct Effect (CDE). We develop regularizations to tackle classical fairness measures and present a causal regularization that satisfies our new fairness definition by removing the impact of unprivileged group variables on the model outcomes as measured by the CDE. These regularizations are applicable to any model trained using by iteratively minimizing a loss through differentiation. We demonstrate our approaches using both gradient boosting and logistic regression on: a synthetic dataset, the UCI Adult (Census) Dataset, and a real-world credit-risk dataset. Our results were found to mitigate unfairness from the predictions with small reductions in model performance.

Bayesian optimization (BO) is a class of sample-efficient global optimization methods, where a probabilistic model conditioned on previous observations is used to determine future evaluations via the optimization of an acquisition function. Most acquisition functions are myopic, meaning that they only consider the impact of the next function evaluation. Non-myopic acquisition functions consider the impact of the next $h$ function evaluations and are typically computed through rollout, in which $h$ steps of BO are simulated. These rollout acquisition functions are defined as $h$-dimensional integrals, and are expensive to compute and optimize. We show that a combination of quasi-Monte Carlo, common random numbers, and control variates significantly reduce the computational burden of rollout. We then formulate a policy-search based approach that removes the need to optimize the rollout acquisition function. Finally, we discuss the qualitative behavior of rollout policies in the setting of multi-modal objectives and model error.

Mean field theory provides an effective way of scaling multiagent reinforcement learning algorithms to environments with many agents that can be abstracted by a virtual mean agent. In this paper, we extend mean field multiagent algorithms to multiple types. The types enable the relaxation of a core assumption in mean field games, which is that all agents in the environment are playing almost similar strategies and have the same goal. We conduct experiments on three different testbeds for the field of many agent reinforcement learning, based on the standard MAgents framework. We consider two different kinds of mean field games: a) Games where agents belong to predefined types that are known a priori and b) Games where the type of each agent is unknown and therefore must be learned based on observations. We introduce new algorithms for each type of game and demonstrate their superior performance over state of the art algorithms that assume that all agents belong to the same type and other baseline algorithms in the MAgent framework.

A Tesla driver killed in a Silicon Valley crash was playing a video game on his smartphone at the time of his fatal crash, investigators said on Tuesday. The National Transport and Safety Board (NTSB) investigation found that Walter Huang, a 38-year-old Apple software engineer and game developer, made no attempts to stop his vehicle as it sped towards a crash barrier before the 2018 crash. Huang's Tesla Model X was in "Autopilot" mode and traveling at about 70 miles per hour when it crashed into a safety barrier and was struck by two other vehicles. He died in hospital from his injuries. "If you own a car with partial automation, do you not own a self-driving car. So don't pretend that you do," said the NTSB chairman, Robert Sumwalt.

We're excited to announce that Gartner has recognized TIBCO Software as a Leader in the 2020 Magic Quadrant for Data Science and Machine Learning Platforms for the 2nd year in a row! We believe Gartner's evaluation validates the innovative digital transformation success our customers have realized across many industries--including financial services, telecom, healthcare, retail, travel and logistics, manufacturing, energy and utilities and the pharmaceuticals. From reporting and modern BI to descriptive and predictive analytics to streaming analytics, TIBCO can help you compete and win. Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research organization and should not be construed as statements of fact.

Atlas 900 stood out with its world-leading AI computing power, ultimate heat dissipation system, and best-in-class cluster network. Atlas 900 accelerates global basic AI research and quickly brings AI applications to industries to advance the AI era with unparalleled AI computing power. Innovative technology has propelled the mobile industry far beyond the wildest expectations of early tech pioneers. GSMA awards the GLOMO Award – Tech of the Future Award to recognize technology that is ahead of its time and reshapes the world. Atlas 900 is the world's fastest AI training cluster.

HeartVista, a pioneer in AI-assisted MRI solutions, announced the formation of its Medical and Scientific Advisory Board, including notable thought leaders from Stanford University and the University of Wisconsin. "The past year was an inflection point for HeartVista, which was full of significant milestones as we received FDA 510(k) clearance for our AI-assisted One Click Cardiac Package," said Itamar Kandel, CEO of HeartVista. "This year, we will continue to progress our MRI software platform and expand its use across additional radiology centers within the US and globally. Our Medical and Scientific Advisory Board will provide strategic direction to our leadership team, enabling us to continue advancing the MRI field." "Automated, AI-driven prescription, as pioneered by HeartVista will change the way we perform advanced MRI exams, by dramatically reducing exam time, standardizing acquisitions, reducing error and rework, and ultimately improve the patient experience," said Dr. Scott Reeder, Vice Chair of Research and Chief of MRI, University of Wisconsin School of Medicine.

Several studies have demonstrated the need to significantly increase the world's food production by 2050. However, there is limited amount of additional arable land, and water levels have also been receding. Although technology could help the farmer, its adoption is limited because the farms usually do not have power, or Internet connectivity, and the farmers are typically not technology savvy. We are working towards an end-to-end approach, from sensors to the cloud, to solve the problem. Our goal is to enable data-driven farming.

Efforts to benchmark computer systems, known as MLPerf, are essential to measure the expanding world of artificial intelligence silicon, according to Jeff Dean, head of AI efforts at Google, but the benchmarks will also have to evolve to better reflect real-world concerns. If computer systems are to evolve to handle ever larger machine learning models, a standard way to compare the effectiveness of those systems is essential, according to Google head of AI, Jeff Dean. But that system of measurement itself must evolve over time, he said. "I think the MLPerf benchmark suite is actually going to be very effective," said Dean, in an interview with ZDNet, last week, referring to the consortium of commercial and academic organizations known as MLPerf, founded within the last few years. The MLPerf group have formulated test suites that measure how different systems do on various AI tasks such as the number of image "convolutions" per second.

Just-released proposals out of Europe that call for new rules to regulate high-risk artificial intelligence systems provide another marker for U.S. insurers and regulators as they consider the opportunities and risks of this evolving technology, industry experts say. The proposals and accompanying data strategy unveiled Feb. 19 are part of the EU's broader digital strategy aimed at setting global standards on technological development that put people first. In its report, the European Commission says that while artificial intelligence can bring advances by tackling climate change and making production more efficient, it also "entails a number of risks, such as opaque decision-making, gender-based or other kinds of discrimination, intrusion in our private lives, or being used for criminal purposes." Jon Godfread, insurance commissioner for North Dakota, said the policy document is "another fencepost and guideline that we can all take a look at" as international discussions on how to regulate artificial intelligence continue to develop. The EU's risk-based regulatory approach outlined in the report says clear rules are needed for high-risk artificial intelligence systems in recruitment, health care, transport, energy and law enforcement so that they are "transparent, traceable and guarantee human oversight."