Clustering is a fundamental unsupervised learning approach. Many clustering algorithms -- such as $k$-means -- rely on the euclidean distance as a similarity measure, which is often not the most relevant metric for high dimensional data such as images. Learning a lower-dimensional embedding that can better reflect the geometry of the dataset is therefore instrumental for performance. We propose a new approach for this task where the embedding is performed by a differentiable model such as a deep neural network. By rewriting the $k$-means clustering algorithm as an optimal transport task, and adding an entropic regularization, we derive a fully differentiable loss function that can be minimized with respect to both the embedding parameters and the cluster parameters via stochastic gradient descent. We show that this new formulation generalizes a recently proposed state-of-the-art method based on soft-$k$-means by adding constraints on the cluster sizes. Empirical evaluations on image classification benchmarks suggest that compared to state-of-the-art methods, our optimal transport-based approach provide better unsupervised accuracy and does not require a pre-training phase.

This paper proposes a new evaluation metric and boosting method for weight separability in neural network design. In contrast to general visual recognition methods designed to encourage both intra-class compactness and inter-class separability of latent features, we focus on estimating linear independence of column vectors in weight matrix and improving the separability of weight vectors. To this end, we propose an evaluation metric for weight separability based on semi-orthogonality of a matrix and Frobenius distance, and the feed-backward reconstruction loss which explicitly encourages weight separability between the column vectors in the weight matrix. The experimental results on image classification and face recognition demonstrate that the weight separability boosting via minimization of feed-backward reconstruction loss can improve the visual recognition performance, hence universally boosting the performance on various visual recognition tasks.

A recent trend of fair machine learning is to define fairness as causality-based notions which concern the causal connection between protected attributes and decisions. However, one common challenge of all causality-based fairness notions is identifiability, i.e., whether they can be uniquely measured from observational data, which is a critical barrier to applying these notions to real-world situations. In this paper, we develop a framework for measuring different causality-based fairness. We propose a unified definition that covers most of previous causality-based fairness notions, namely the path-specific counterfactual fairness (PC fairness). Based on that, we propose a general method in the form of a constrained optimization problem for bounding the path-specific counterfactual fairness under all unidentifiable situations. Experiments on synthetic and real-world datasets show the correctness and effectiveness of our method.

The Artificial Intelligence Robots Market report is a complete overview of the market, covering various aspects product definition, segmentation based on various parameters, and the prevailing vendor landscape. Analysis and discussion of important industry trends, market size, market share estimates are mentioned in the report. Artificial Intelligence Robots Market report includes historic data, present market trends, environment, technological innovation, upcoming technologies and the technical progress in the related industry. The Global Artificial Intelligence Robots Market accounted for USD 3.0 billion in 2017 and is projected to grow at a CAGR of 30.1% forecast to 2025. Some of the major countries covered in this report are U.S., Canada, Germany, France, U.K., Netherlands, Switzerland, Turkey, Russia, China, India, South Korea, Japan, Australia, Singapore, Saudi Arabia, South Africa and Brazil among others.

BMO Capital Markets is a trade name used by BMO Financial Group for the wholesale banking businesses of Bank of Montreal, BMO Harris Bank N.A. (member FDIC), Bank of Montreal Europe p.l.c, and Bank of Montreal (China) Co. Ltd and the institutional broker dealer businesses of BMO Capital Markets Corp.

By Howard Solomon A new report highlights a growing fear among Canadians that's tied to the rapid advancement of artificial intelligence. An online study conducted by Environics Research Group revealed that 77 per cent of Canadians are concerned that AI is advancing too quickly to properly understand its potential risks. The survey of 1,200 Canadians was sponsored by TD Bank back in May, and also indicated a growing concern around biases in how the technology is developed. Additionally, sixty per cent of Canadians worry about a lack of diversity in the growing field of AI. The results don't shock Tomi Poutanen, chief AI officer for TD and co-founder of Layer 6, but he said they do signal a growing awareness of AI's transformative capabilities, and people are looking to banks to validate its adoption.

EPAM Systems, Inc. (NYSE: EPAM), a leading global provider of digital platform engineering and software development services, today announced that it has been named a Diamond Global Business Partner of UiPath, an enterprise Robotic Process Automation (RPA) software company. EPAM and UiPath's partnership will enable its joint customers to increase efficiencies and improve customer experience by leveraging intelligent automation (IA) solutions and UiPath's RPA platform. Forrester recently predicted that, in 2019, automation would become the tip of the digital transformation spear. While early automation implementations focused on cost optimization, this new wave will achieve multiple goals, including driving both customer and employee experience, changing the nature of work, and even empowering the next generation of startup companies. With more than 10 years of business process management, robotics and cognitive expertise, EPAM has over 100 certified UiPath advanced developers as part of its team of more than 700 machine learning and RPA engineers.

One of the technologies we are seeing being trialled and deployed in airports is robotic assistants. The humanoid robots are positioned around the airport terminal assisting passengers with queries and information. By making use of Artificial Intelligence (AI) and Machine Learning, the robots can process large amounts of data, with real-time updates to enable them to provide the latest information to passengers. This technology is starting to be used in some select airports but for different functions. Munich Airport in Germany is using robotic assistants primarily for information.

Artificial Intelligence is poised to disrupt many industries, but education arena has not typically been at the forefront of such conversations. If it has been included at all, the narrative has been in a more abstract manner than actual application. And even though several companies such as Carnegie Learning and Content Technologies, Inc have taken either more adult learning approaches or those that are deeply rooted in tech, the space is still anyone's game with new trends to be developed for Gen Z. The industry is an important one not only for its ability to generate an entirely new level of learning but also because of the very real business opportunity in the space. Indeed, the artificial intelligence in education size is forecasted at a market size worth $6 billion dollars by 2024.

Transferring knowledge across many streaming processes remains an uncharted territory in the existing literature and features unique characteristics: no labelled instance of the target domain, covariate shift of source and target domain, different period of drifts in the source and target domains. Autonomous transfer learning (ATL) is proposed in this paper as a flexible deep learning approach for the online unsupervised transfer learning problem across many streaming processes. ATL offers an online domain adaptation strategy via the generative and discriminative phases coupled with the KL divergence based optimization strategy to produce a domain invariant network while putting forward an elastic network structure. It automatically evolves its network structure from scratch with/without the presence of ground truth to overcome independent concept drifts in the source and target domain. The rigorous numerical evaluation has been conducted along with a comparison against recently published works. ATL demonstrates improved performance while showing significantly faster training speed than its counterparts.