Machine learning is the common basis for modern artificial intelligence, using large amounts of data to build AI models that recognize patterns and make predictions when presented with new information. A new Seattle startup led by a former Microsoft distinguished engineer uses a different approach: machine teaching. "It's not extracting knowledge from data; it's extracting knowledge from the person," explained Patrice Simard, CEO and co-founder of Intelus, who oversaw Microsoft research groups in areas including machine learning, databases, graphics, vision and cryptography in more than 20 years at the Redmond company. Intelus emerged from stealth mode Tuesday to launch an open beta of its new machine teaching platform, Duet, which offers a graphical user interface to create AI models from unstructured data without writing code or requiring advanced data science tools. The models can then be used to classify and extract data from text.

CORVALLIS, Ore. – A new study by the Oregon State University College of Engineering shows that machine learning techniques can offer powerful new tools for advancing personalized medicine, care that optimizes outcomes for individual patients based on unique aspects of their biology and disease features. The research with machine learning, a branch of artificial intelligence in which computer systems use algorithms and statistical models to look for trends in data, tackles long-unsolvable problems in biological systems at the cellular level, said Oregon State's Brian D. Wood, who conducted the study with then OSU Ph.D. student Ehsan Taghizadeh and Helen M. Byrne of the University of Oxford. "Those systems tend to have high complexity – first because of the vast number of individual cells and second, because of the highly nonlinear way in which cells can behave," said Wood, a professor of environmental engineering. "Nonlinear systems present a challenge for upscaling methods, which is the primary means by which researchers can accurately model biological systems at the larger scales that are often the most relevant." A linear system in science or mathematics means any change to the system's input results in a proportional change to the output; a linear equation, for example, might describe a slope that gains 2 feet vertically for every foot of horizontal distance.

Have you ever spent hours searching for a suitable dataset for your data science project? It can get pretty daunting, right? Whether you are a student or a professional looking for high-quality datasets for machine learning or data analysis projects--we've got you covered! In today's article, we will share with you a comprehensive list of 65 open machine learning datasets that you can access for free. We will regularly update this list, so feel free to suggest datasets you are using and we will make sure to add them. "Where can I get free datasets for machine learning?" Here's the list of the best open dataset finders that you can use to browse through a wide variety of niche-specific datasets for your data science projects.

"I prefer to stay out of politics." Those were Elon Musk's words when the tech exec was forced to respond to a claim by Texas governor Greg Abbott that he supported the state's anti-abortion laws. If so, Mr Musk has a funny way of showing it. Over his decade-plus of public fame as the chief executive of Tesla and SpaceX, the South-African-born tycoon has attacked everyone and everything, from Donald Trump and Bernie Sanders through individual regulatory officials to Covid rules, trade unions, and "pronouns". On Monday, he hammered US president Joe Biden's flagship infrastructure and social spending bills for granting unnecessary subsidies to the electric car industry and increasing the "insane" federal budget deficit.

Humans are working artificial intelligence programs (AI) into business, government and daily life. Like with any new tool or technology, we start to see the initial technology flaws the more we are exposed to it. So we are now in the midst of a moment where AI is under the microscope, with policy makers picking apart AI contributions and demanding that AI meet high standards of performance and social consequence. This is a healthy process. Society should always examine impactful tools and push for the tools to work better.

Artificial intelligence (AI) can be used to strengthen cybersecurity by detecting cyber threats more quickly and effectively. Imagine, for instance, if embedded malware meant to disrupt a hospital system's operation was detected within minutes of its installation, rather than after it disrupted access to electronic medical health records, delaying medical treatments? In such a scenario, the use of AI to detect a cyber-attack could literally mean life or death. Despite these benefits, the use of AI also introduces its own set of cybersecurity vulnerabilities that bad actors can exploit. For example, it can amplify the exploitation of vulnerabilities across digital systems.

The UK government's Centre for Data Ethics and Innovation (CDEI) has published a "roadmap" designed to create an AI assurance industry to support the introduction of automated analysis, decision making, and processes. The move is one of several government initiatives planned to help shape local development and use of AI – an industry that attracted £2.5bn investment in 2019 – but it raises as many questions as it answers. Part of the Department for Digital, Culture, Media & Sport (DCMS), the CDEI said by "verifying that AI systems are effective, trustworthy and compliant, AI assurance services will drive a step-change in adoption, enabling the UK to realise the full potential of AI and develop a competitive edge." Launching the move, DCMS minister Chris Philp said: "The roadmap sets out the steps needed to grow a mature, world-class AI assurance industry. AI assurance services will become a key part of the toolkit available to ensure effective, pro-innovation governance of AI." How that governance will take shape is, as yet, a bit fuzzy while the industry waits on proposals for AI legislation in the forthcoming White Paper on governance and regulation.

Recent studies have shown that robustness to adversarial attacks can be transferred across networks. In other words, we can make a weak model more robust with the help of a strong teacher model. We ask if instead of learning from a static teacher, can models "learn together" and "teach each other" to achieve better robustness? In this paper, we study how interactions among models affect robustness via knowledge distillation. We propose mutual adversarial training (MAT), in which multiple models are trained together and share the knowledge of adversarial examples to achieve improved robustness. MAT allows robust models to explore a larger space of adversarial samples, and find more robust feature spaces and decision boundaries. Through extensive experiments on CIFAR-10 and CIFAR-100, we demonstrate that MAT can effectively improve model robustness and outperform state-of-the-art methods under white-box attacks, bringing $\sim$8% accuracy gain to vanilla adversarial training (AT) under PGD-100 attacks. In addition, we show that MAT can also mitigate the robustness trade-off among different perturbation types, bringing as much as 13.1% accuracy gain to AT baselines against the union of $l_\infty$, $l_2$ and $l_1$ attacks. These results show the superiority of the proposed method and demonstrate that collaborative learning is an effective strategy for designing robust models.

A model's interpretability is essential to many practical applications such as clinical decision support systems. In this paper, a novel interpretable machine learning method is presented, which can model the relationship between input variables and responses in humanly understandable rules. The method is built by applying tropical geometry to fuzzy inference systems, wherein variable encoding functions and salient rules can be discovered by supervised learning. Experiments using synthetic datasets were conducted to investigate the performance and capacity of the proposed algorithm in classification and rule discovery. Furthermore, the proposed method was applied to a clinical application that identified heart failure patients that would benefit from advanced therapies such as heart transplant or durable mechanical circulatory support. Experimental results show that the proposed network achieved great performance on the classification tasks. In addition to learning humanly understandable rules from the dataset, existing fuzzy domain knowledge can be easily transferred into the network and used to facilitate model training. From our results, the proposed model and the ability of learning existing domain knowledge can significantly improve the model generalizability. The characteristics of the proposed network make it promising in applications requiring model reliability and justification.

Autonomous vision-based spaceborne navigation is an enabling technology for future on-orbit servicing and space logistics missions. While computer vision in general has benefited from Machine Learning (ML), training and validating spaceborne ML models are extremely challenging due to the impracticality of acquiring a large-scale labeled dataset of images of the intended target in the space environment. Existing datasets, such as Spacecraft PosE Estimation Dataset (SPEED), have so far mostly relied on synthetic images for both training and validation, which are easy to mass-produce but fail to resemble the visual features and illumination variability inherent to the target spaceborne images. In order to bridge the gap between the current practices and the intended applications in future space missions, this paper introduces SPEED+: the next generation spacecraft pose estimation dataset with specific emphasis on domain gap. In addition to 60,000 synthetic images for training, SPEED+ includes 9,531 hardware-in-the-loop images of a spacecraft mockup model captured from the Testbed for Rendezvous and Optical Navigation (TRON) facility. TRON is a first-of-a-kind robotic testbed capable of capturing an arbitrary number of target images with accurate and maximally diverse pose labels and high-fidelity spaceborne illumination conditions. SPEED+ is used in the second international Satellite Pose Estimation Challenge co-hosted by SLAB and the Advanced Concepts Team of the European Space Agency to evaluate and compare the robustness of spaceborne ML models trained on synthetic images.