Mislabeled examples are ubiquitous in real-world machine learning datasets, advocating the development of techniques for automatic detection. We show that most mislabeled detection methods can be viewed as probing trained machine learning models using a few core principles. We formalize a modular framework that encompasses these methods, parameterized by only 4 building blocks, as well as a Python library that demonstrates that these principles can actually be implemented. The focus is on classifier-agnostic concepts, with an emphasis on adapting methods developed for deep learning models to non-deep classifiers for tabular data. We benchmark existing methods on (artificial) Completely At Random (NCAR) as well as (realistic) Not At Random (NNAR) labeling noise from a variety of tasks with imperfect labeling rules. This benchmark provides new insights as well as limitations of existing methods in this setup.

Density based representations of atomic environments that are invariant under Euclidean symmetries have become a widely used tool in the machine learning of interatomic potentials, broader data-driven atomistic modelling and the visualisation and analysis of materials datasets.The standard mechanism used to incorporate chemical element information is to create separate densities for each element and form tensor products between them. This leads to a steep scaling in the size of the representation as the number of elements increases. Graph neural networks, which do not explicitly use density representations, escape this scaling by mapping the chemical element information into a fixed dimensional space in a learnable way. We recast this approach as tensor factorisation by exploiting the tensor structure of standard neighbour density based descriptors. In doing so, we form compact tensor-reduced representations whose size does not depend on the number of chemical elements, but remain systematically convergeable and are therefore applicable to a wide range of data analysis and regression tasks.

Recently, the application of machine learning models has gained momentum in natural sciences and engineering, which is a natural fit due to the abundance of data in these fields. However, the modeling of physical processes from simulation data without first principle solutions remains difficult. Here, we present a Graph Neural Networks approach towards accurate modeling of complex 3D granular flow simulation processes created by the discrete element method LIGGGHTS and concentrate on simulations of physical systems found in real world applications like rotating drums and hoppers. We discuss how to implement Graph Neural Networks that deal with 3D objects, boundary conditions, particle - particle, and particle - boundary interactions such that an accurate modeling of relevant physical quantities is made possible. Finally, we compare the machine learning based trajectories to LIGGGHTS trajectories in terms of particle flows and mixing entropies.

Scientists aim to extract simplicity from observations of the complex world. An important component of this process is the exploration of data in search of trends. In practice, however, this tends to be more of an art than a science. Among all trends existing in the natural world, one-dimensional trends, often called sequences, are of particular interest as they provide insights into simple phenomena. However, some are challenging to detect as they may be expressed in complex manners. We present the Sequencer, an algorithm designed to generically identify the main trend in a dataset. It does so by constructing graphs describing the similarities between pairs of observations, computed with a set of metrics and scales. Using the fact that continuous trends lead to more elongated graphs, the algorithm can identify which aspects of the data are relevant in establishing a global sequence. Such an approach can be used beyond the proposed algorithm and can optimize the parameters of any dimensionality reduction technique. We demonstrate the power of the Sequencer using real-world data from astronomy, geology as well as images from the natural world. We show that, in a number of cases, it outperforms the popular t-SNE and UMAP dimensionality reduction techniques. This approach to exploratory data analysis, which does not rely on training nor tuning of any parameter, has the potential to enable discoveries in a wide range of scientific domains. The source code is available on github and we provide an online interface at \url{http://sequencer.org}.

Five Dutch companies Ahold Delhaize, ING, KLM, NS and Philips aim to further boost the AI ecosystem in the Netherlands by accelerating and promoting the development of AI technology and nurturing AI talent in the country. This effort will add educational capacity, foster the development of the AI community in the Netherlands and reiterate the position of the Netherlands as a competitive and relevant global AI hub. The goal of Kickstart AI, is to bridge the AI gap between the Netherlands and other countries, like the UK, the US and China, that have made notable progress in this area. In order to keep the country's position as a pioneer and inventor of technologies, the Dutch government, companies, organizations and universities have ground to cover in terms of structural investments and availability of global AI talent. The five companies "kickstarting" AI are, for the first time, uniting forces in this kind of joint initiative and taking highly needed decisive action.

Five Dutch companies Ahold Delhaize, ING, KLM, NS and Philips aim to further boost the AI ecosystem in the Netherlands by accelerating and promoting the development of AI technology and nurturing AI talent in the country. This effort will add educational capacity, foster the development of the AI community in the Netherlands and reiterate the position of the Netherlands as a competitive and relevant global AI hub. The goal of Kickstart AI, is to bridge the AI gap between the Netherlands and other countries, like the UK, the US and China, that have made notable progress in this area. In order to keep the country's position as a pioneer and inventor of technologies, the Dutch government, companies, organizations and universities have ground to cover in terms of structural investments and availability of global AI talent. The five companies "kickstarting" AI are, for the first time, uniting forces in this kind of joint initiative and taking highly needed decisive action.

"The most beautiful part of the whole thing is when we started the project we didn't really know each other, but slowly we've grown up together." It's a remark that you might expect to hear in a scene from one of your favourite movies, or perhaps in a confession from a colleague on a team-building day. But it's actually how AI tech startup Aiir Innovation's CEO Bart Vredebregt explains how five AI students and an AI professor came together to launch a business that could revolutionise the maintenance industry. It all started in 2015, when Dutch airline KLM invited students on the University of Amsterdam AI master's degree programme to work on a project to automate the inspection of their jet engines. When done manually, this inspection requires an engineer to put a camera called a borescope into the engine after removing the aircraft's wing.

After years of close cooperation, the Boston Consulting Group and KLM Royal Dutch Airlines have agreed to what they describe as a pioneering artificial intelligence partnership that could "revolutionise global airline operations". It is the first time in the Dutch airline's history that it has collaborated with a leading management consultancy to launch an entirely new service. The jointly-developed artificial intelligence (AI) system will digitise KLM's entire commercial aviation process and leverage advanced machine learning technology to streamline operations "in an unprecedented manner". With these and other state-of-the-art analytic and organisational tools, the suite of integrated solutions will be able to minimise daily disruptions caused by human error and unforeseen events. Combining AI, machine learning, and advanced analytics with a joint force of KLM Operational staff and data scientists, engineers and developers from BCG Gamma (the data science consulting arm of BCG), the technology can help optimise airline processes, from front-office to back-office.

While a data-based world allows us to build services and products, it also requires complicated processes of acquisition that can be tedious if not overwhelming. This is why businesses are increasingly implementing artificial intelligence to simplify their data-based services. Here's how 5 brands are turning to chatbots and voice programs to allow users to easily input information and receive personalized instructions for complex or unfamiliar processes, resulting in a customer onboarding journey with minimal friction. Bing Bing created chatbots embedded into search results to allow customers to interact with businesses to get their basic questions answered. The bot, which is available across channels and platforms, answers users questions based on pre-answered questions by business owners, and if there's a question it can't answer, the bot will refer the user to a phone number.

Dutch carrier KLM, already an airline industry leader in the use of artificial intelligence (AI) to field customer service inquiries through social media channels, can now deal with many social media interactions without a live agent. The enhancement allows it to answer more questions in a shorter period of time. "This is exactly what the customer needs," said Air France-KLM senior vice president of digital Pieter Groeneveld. KLM said its team of 250 social media service agents engage in approximately 30,000 conversations each week, double the volume they were handling just 13 months ago. On average, conversations consist of five or six questions and answers.