We consider the problem of optimizing the distribution operations of a hub using drones to deliver medical supplies to different geographic regions. Drones are an innovative method with many benefits including low-contact delivery thereby reducing the spread of pandemic and vaccine-preventable diseases. While we focus on medical supply delivery for this work, it is applicable to drone delivery for many other applications, including food, postal items, and e-commerce delivery. In this paper, our goal is to address drone delivery challenges by optimizing the distribution operations at a drone hub that dispatch drones to different geographic locations generating stochastic demands for medical supplies. By considering different geographic locations, we consider different classes of demand that require different flight ranges, which is directly related to the amount of charge held in a drone battery. We classify the stochastic demands based on their distance from the drone hub, use a Markov decision process to model the problem, and perform computational tests using realistic data representing a prominent drone delivery company. We solve the problem using a reinforcement learning method and show its high performance compared with the exact solution found using dynamic programming. Finally, we analyze the results and provide insights for managing the drone hub operations.

The advent of contextual word embeddings -- representations of words which incorporate semantic and syntactic information from their context -- has led to tremendous improvements on a wide variety of NLP tasks. However, recent contextual models have prohibitively high computational cost in many use-cases and are often hard to interpret. In this work, we demonstrate that our proposed distillation method, which is a simple extension of CBOW-based training, allows to significantly improve computational efficiency of NLP applications, while outperforming the quality of existing static embeddings trained from scratch as well as those distilled from previously proposed methods. As a side-effect, our approach also allows a fair comparison of both contextual and static embeddings via standard lexical evaluation tasks.

We developed Distilled Graph Attention Policy Networks (DGAPNs), a curiosity-driven reinforcement learning model to generate novel graph-structured chemical representations that optimize user-defined objectives by efficiently navigating a physically constrained domain. The framework is examined on the task of generating molecules that are designed to bind, noncovalently, to functional sites of SARS-CoV-2 proteins. We present a spatial Graph Attention Network (sGAT) that leverages self-attention over both node and edge attributes as well as encoding spatial structure -- this capability is of considerable interest in areas such as molecular and synthetic biology and drug discovery. An attentional policy network is then introduced to learn decision rules for a dynamic, fragment-based chemical environment, and state-of-the-art policy gradient techniques are employed to train the network with enhanced stability. Exploration is efficiently encouraged by incorporating innovation reward bonuses learned and proposed by random network distillation. In experiments, our framework achieved outstanding results compared to state-of-the-art algorithms, while increasing the diversity of proposed molecules and reducing the complexity of paths to chemical synthesis.

Self-supervised representation learning has shown remarkable success in a number of domains. A common practice is to perform data augmentation via hand-crafted transformations intended to leave the semantics of the data invariant. We seek to understand the empirical success of this approach from a theoretical perspective. We formulate the augmentation process as a latent variable model by postulating a partition of the latent representation into a content component, which is assumed invariant to augmentation, and a style component, which is allowed to change. Unlike prior work on disentanglement and independent component analysis, we allow for both nontrivial statistical and causal dependencies in the latent space. We study the identifiability of the latent representation based on pairs of views of the observations and prove sufficient conditions that allow us to identify the invariant content partition up to an invertible mapping in both generative and discriminative settings. We find numerical simulations with dependent latent variables are consistent with our theory. Lastly, we introduce Causal3DIdent, a dataset of high-dimensional, visually complex images with rich causal dependencies, which we use to study the effect of data augmentations performed in practice.

Network regression models, where the outcome comprises the valued edge in a network and the predictors are actor or dyad-level covariates, are used extensively in the social and biological sciences. Valid inference relies on accurately modeling the residual dependencies among the relations. Frequently homogeneity assumptions are placed on the errors which are commonly incorrect and ignore critical, natural clustering of the actors. In this work, we present a novel regression modeling framework that models the errors as resulting from a community-based dependence structure and exploits the subsequent exchangeability properties of the error distribution to obtain parsimonious standard errors for regression parameters.

This paper introduces a generative model equivariant to Euclidean symmetries: E(n) Equivariant Normalizing Flows (E-NFs). To construct E-NFs, we take the discriminative E(n) graph neural networks and integrate them as a differential equation to obtain an invertible equivariant function: a continuous-time normalizing flow. We demonstrate that E-NFs considerably outperform baselines and existing methods from the literature on particle systems such as DW4 and LJ13, and on molecules from QM9 in terms of log-likelihood. To the best of our knowledge, this is the first flow that jointly generates molecule features and positions in 3D.

A Moscow-based startup called Neiry is developing "human-friendly neural interfaces" applicable to a variety of daily tasks -- from gaming and entertainment, to medical therapy, to training, to business. Supported by Russian state innovation programs, this startup believes it can "improve and restore human abilities and functions that were previously unattainable," and contribute to "the acceleration of computer-integrated human evolution." Neiry's approach combines headsets with an advanced neural control approach. Its technology is non-invasive in the sense that no device needs to be implanted: "You simply wear it on your head just like you put on a watch or a sports bracelet on your hand," says the company. "Our neural networks are trained in such a way that the user does not need to learn how to work with the neuro-headset, unlike other 99% of similar devices. As a result, users can perform "up to eight different actions per unit of a digital image, while 99% of similar devices barely perform two or three actions." The startup claims its solution is equally compatible with eye tracking, electrodermal activity (EDA), and speech recognition technologies. Combined with brain-computer interfaces, the latter can multiply the number of commands when using the neural interface. Neiry is now developing an infrastructure for automated collection, processing and storage of large amounts or neural data. "Thanks to unique algorithms, the brain's bioelectric activity can be analyzed, enabling users of the newest edtech applications and games to control the related parameters of brain signals.

The medical industry is sitting on a huge trove of data, but in many cases it can be a challenge to realize the value of it because that data is unstructured and in disparate places. Today, a startup called Mendel, which has built an AI platform both to ingest and bring order to that body of information, is announcing $18 million in funding to continue its growth and to build out what it describes as a "clinical data marketplace" for people not just to organize, but also to share and exchange that data for research purposes. It's also going to be using the funding to hire more talent -- technical and support -- for its two offices, in San Jose, CA and Cairo, Egypt. The Series A round is being led by DCM, with OliveTree and MTVLP, and previous backers Launch Capital, SOSV, Bootstrap Labs and Chairman of UCSF Health Hub Mark Goldstein also participating. The funding comes on the heels of what Mendel says is a surge of interest among research and pharmaceutical companies in sourcing better data to gain a better understanding of longer-term patient care and progress, in particular across wider groups of users, not just at a time when it has been more challenging to observe people and run trials, but in light of the understanding that using AI to leverage much bigger data sets can produce better insights.

From a ship that has been designed to tell you what maintenance it needs and when, to an intelligent journey planner for global goods transport. The three universities in Zuid-Holland are abuzz with AI research in the field of ports and maritime. "The big challenges facing ports are accessibility, sustainability and finding the right employees," says Rudy Negenborn, Professor of Multi-Machine Operations & Logistics in Delft. "In a busy port, you have to optimise your planning to avoid delays, congestion and unnecessary emissions. This doesn't just require solutions to technical challenges: a solution can only be implemented if an organisation wants and has the right infrastructure for this."

In brief Machines are like humans – they're lazy. When given the chance to take the easy route to complete an easy task, they will. Academics at the University of Washington found that algorithms trained to diagnose COVID-19 from chest X-rays often look at secondary features, such as a patient's age, rather than focusing on the images themselves – something known as shortcut learning. "A physician would generally expect a finding of COVID-19 from an X-ray to be based on specific patterns in the image that reflect disease processes," said Alex DeGrave, a medical science student at the American university and co-author of a paper published this week in Nature Intelligence. "But, rather than relying on those patterns, a system using shortcut learning might, for example, judge that someone is elderly and, thus, infer that they are more likely to have the disease because it is more common in older patients. The shortcut is not wrong per se, but the association is unexpected and not transparent. And, that could lead to an inappropriate diagnosis."