Predicting agents' future trajectories plays a crucial role in modern AI systems, yet it is challenging due to intricate interactions exhibited in multi-agent systems, especially when it comes to collision avoidance. To address this challenge, we propose to learn congestion patterns as contextual cues explicitly and devise a novel "Sense--Learn--Reason--Predict" framework by exploiting advantages of three different doctrines of thought, which yields the following desirable benefits: (i) Representing congestion as contextual cues via latent factors subsumes the concept of social force commonly used in physics-based approaches and implicitly encodes the distance as a cost, similar to the way a planning-based method models the environment. (ii) By decomposing the learning phases into two stages, a "student" can learn contextual cues from a "teacher" while generating collision-free trajectories. To make the framework computationally tractable, we formulate it as an optimization problem and derive an upper bound by leveraging the variational parametrization. In experiments, we demonstrate that the proposed model is able to generate collision-free trajectory predictions in a synthetic dataset designed for collision avoidance evaluation and remains competitive on the commonly used NGSIM US-101 highway dataset.

We introduce a visually-guided and physics-driven task-and-motion planning benchmark, which we call the ThreeDWorld Transport Challenge. In this challenge, an embodied agent equipped with two 9-DOF articulated arms is spawned randomly in a simulated physical home environment. The agent is required to find a small set of objects scattered around the house, pick them up, and transport them to a desired final location. We also position containers around the house that can be used as tools to assist with transporting objects efficiently. To complete the task, an embodied agent must plan a sequence of actions to change the state of a large number of objects in the face of realistic physical constraints. We build this benchmark challenge using the ThreeDWorld simulation: a virtual 3D environment where all objects respond to physics, and where can be controlled using fully physics-driven navigation and interaction API. We evaluate several existing agents on this benchmark. Experimental results suggest that: 1) a pure RL model struggles on this challenge; 2) hierarchical planning-based agents can transport some objects but still far from solving this task. We anticipate that this benchmark will empower researchers to develop more intelligent physics-driven robots for the physical world.

Predicting accurate future trajectories of multiple agents is essential for autonomous systems, but is challenging due to the complex agent interaction and the uncertainty in each agent's future behavior. Forecasting multi-agent trajectories requires modeling two key dimensions: (1) time dimension, where we model the influence of past agent states over future states; (2) social dimension, where we model how the state of each agent affects others. Most prior methods model these two dimensions separately; e.g., first using a temporal model to summarize features over time for each agent independently and then modeling the interaction of the summarized features with a social model. This approach is suboptimal since independent feature encoding over either the time or social dimension can result in a loss of information. Instead, we would prefer a method that allows an agent's state at one time to directly affect another agent's state at a future time. To this end, we propose a new Transformer, AgentFormer, that jointly models the time and social dimensions. The model leverages a sequence representation of multi-agent trajectories by flattening trajectory features across time and agents. Since standard attention operations disregard the agent identity of each element in the sequence, AgentFormer uses a novel agent-aware attention mechanism that preserves agent identities by attending to elements of the same agent differently than elements of other agents. Based on AgentFormer, we propose a stochastic multi-agent trajectory prediction model that can attend to features of any agent at any previous timestep when inferring an agent's future position. The latent intent of all agents is also jointly modeled, allowing the stochasticity in one agent's behavior to affect other agents. Our method significantly improves the state of the art on well-established pedestrian and autonomous driving datasets.

The unification of statistical (data-driven) and symbolic (knowledge-driven) methods is widely recognised as one of the key challenges of modern AI. Recent years have seen large number of publications on such hybrid neuro-symbolic AI systems. That rapidly growing literature is highly diverse and mostly empirical, and is lacking a unifying view of the large variety of these hybrid systems. In this paper we analyse a large body of recent literature and we propose a set of modular design patterns for such hybrid, neuro-symbolic systems. We are able to describe the architecture of a very large number of hybrid systems by composing only a small set of elementary patterns as building blocks. The main contributions of this paper are: 1) a taxonomically organised vocabulary to describe both processes and data structures used in hybrid systems; 2) a set of 15+ design patterns for hybrid AI systems, organised in a set of elementary patterns and a set of compositional patterns; 3) an application of these design patterns in two realistic use-cases for hybrid AI systems. Our patterns reveal similarities between systems that were not recognised until now. Finally, our design patterns extend and refine Kautz' earlier attempt at categorising neuro-symbolic architectures.

We study liquid democracy, a collective decision making paradigm that allows voters to transitively delegate their votes, through an algorithmic lens. In our model, there are two alternatives, one correct and one incorrect, and we are interested in the probability that the majority opinion is correct. Our main question is whether there exist delegation mechanisms that are guaranteed to outperform direct voting, in the sense of being always at least as likely, and sometimes more likely, to make a correct decision. Even though we assume that voters can only delegate their votes to better-informed voters, we show that local delegation mechanisms, which only take the local neighborhood of each voter as input (and, arguably, capture the spirit of liquid democracy), cannot provide the foregoing guarantee. By contrast, we design a non-local delegation mechanism that does provably outperform direct voting under mild assumptions about voters.

The paper considers a distributed version of deep reinforcement learning (DRL) for multi-agent decision-making process in the paradigm of federated learning. Since the deep neural network models in federated learning are trained locally and aggregated iteratively through a central server, frequent information exchange incurs a large amount of communication overheads. Besides, due to the heterogeneity of agents, Markov state transition trajectories from different agents are usually unsynchronized within the same time interval, which will further influence the convergence bound of the aggregated deep neural network models. Therefore, it is of vital importance to reasonably evaluate the effectiveness of different optimization methods. Accordingly, this paper proposes a utility function to consider the balance between reducing communication overheads and improving convergence performance. Meanwhile, this paper develops two new optimization methods on top of variation-aware periodic averaging methods: 1) the decay-based method which gradually decreases the weight of the model's local gradients within the progress of local updating, and 2) the consensus-based method which introduces the consensus algorithm into federated learning for the exchange of the model's local gradients. This paper also provides novel convergence guarantees for both developed methods and demonstrates their effectiveness and efficiency through theoretical analysis and numerical simulation results.

Motivation: Several accurate deep learning models have been proposed to predict drug-target affinity (DTA). However, all of these models are black box hence are difficult to interpret and verify its result, and thus risking acceptance. Explanation is necessary to allow the DTA model more trustworthy. Explanation with counterfactual provides human-understandable examples. Most counterfactual explanation methods only operate on single input data, which are in tabular or continuous forms. In contrast, the DTA model has two discrete inputs. It is challenging for the counterfactual generation framework to optimize both discrete inputs at the same time. Results: We propose a multi-agent reinforcement learning framework, Multi-Agent Counterfactual Drug-target binding Affinity (MACDA), to generate counterfactual explanations for the drug-protein complex. Our proposed framework provides human-interpretable counterfactual instances while optimizing both the input drug and target for counterfactual generation at the same time. The result on the Davis dataset shows the advantages of the proposed MACDA framework compared with previous works.

A desirable goal for autonomous agents is to be able to coordinate on the fly with previously unknown teammates. Known as "ad hoc teamwork", enabling such a capability has been receiving increasing attention in the research community. One of the central challenges in ad hoc teamwork is quickly recognizing the current plans of other agents and planning accordingly. In this paper, we focus on the scenario in which teammates can communicate with one another, but only at a cost. Thus, they must carefully balance plan recognition based on observations vs. that based on communication. This paper proposes a new metric for evaluating how similar are two policies that a teammate may be following - the Expected Divergence Point (EDP). We then present a novel planning algorithm for ad hoc teamwork, determining which query to ask and planning accordingly. We demonstrate the effectiveness of this algorithm in a range of increasingly general communication in ad hoc teamwork problems.

The COVID pandemic may be receding, but it has left a mark on across multiple aspects of our lives. From mask mandates to travel restrictions, we chafe at some of the changes – but in the business world the use of artificial intelligence (AI) systems has dramatically expanded in the past year. This was probably inevitable – but AI brought advantages in coping with the pandemic for companies that could make use of it, and the expansion accelerated. AI has found its place in a huge range of applications, at both the front and back end of businesses. It’s prevalent in software management and data systems, as well as in communications, where AI systems filter emails and conduct robochats. And this has not been ignored by Wall Street. Analysts say that plenty of compelling investments can be found within this space. With this in mind, we’ve opened up TipRanks’ database, and pulled two stocks which are stand to benefit from AI technology. Importantly, both have amassed enough bullish calls from analysts to be given “Strong Buy” consensus ratings. Nuance Communications (NUAN) We’ll start with Nuance, a company in the communications software niche. This Massachusetts-based company offers solutions for business clients in the healthcare and customer service industries, with products that enhance speech recognition, telephone call steering systems, automated phone directories, medical transcription, and optical character recognition. It’s a full range of AI-powered, cloud communications software, applied in real time. Nuance’s flagship product, the Dragon Ambient eXperience (DAX) is marketed to the healthcare industry, where it uses AI to automate the paperwork burdens on physician practices and hospitals. This streamlines operations allow doctors more time and resources to spend on patients, and provides greater satisfaction to health care providers and users. The applications of Nuance’s product and solution lines to the current environment is clear: when the pandemic locked down so many people at home, businesses still had to maintain their customer-facing systems, and software automation, based on AI tech, made that possible with fewer personnel. Since the pandemic started last winter, the company seen its shares grow tremendously, up 205% in the last 12 months, far outpacing the overall stock market. The most recent quarterly report, for fiscal Q1, showed quarterly revenues above the forecast at $81.4 million. EPS showed a net loss, as expected, but at 27 cents the loss was a 28% sequential improvement from Q3. The company’s balance sheet is strong, with zero debt, $256 million cash on hand, and a credit facility up to $50 million. The company’s most recent quarterly report, for fiscal Q1, beat the forecasts on both the top and bottom lines. Earnings beat expectations by 11%, coming in at 20 cents per share, while revenues of $345.8 million were a modest 2% above the estimates. As a result, operating cash flow grew 22% year-over-year, to $54.6 million for the quarter. Among the bulls is 5-star analyst Daniel Ives, of Wedbush, who rates NUAN shares an Outperform (i.e. Buy), and his $65 price target implies an upside potential of ~44%. (To watch Ives’ track record, click here) "We believe Nuance overall continues to be laser focused on building a global cloud healthcare and AI driven business with growing ARR and a sustainable revenue/ earnings stream going forward with larger deals in the field as more hospital- wide deployments shift to the cloud are playing out and gaining further momentum based on our checks," Ives opined. The analyst added, "From a valuation/ SOTP perspective, we believe over time the DAX business alone could be worth between $3 billion to $4 billion to NUAN's stock as this AI next generation platform represents a potential paradigm changer for hospitals/healthcare clinics/specialists over the coming years." Ives is no outlier on Nuance, as shown by the unanimous Strong Buy analyst consensus on the stock. Nuance has received 6 recent reviews, and all are to Buy. The shares are trading for $45.20, and the $59.67 average price target suggests a 32% one-year upside. (See NUAN stock analysis on TipRanks) Dynatrace, Inc. (DT) The second AI stock we’ll look at, Dynatrace, is another cloud software company – but Dynatrace’s products are designed to power business data. The company’s AI platform brings intelligent automation to network management and cloud monitoring. DT’s platform allows for cloud automation, business analytics, digital experience, application security, applications and microservices, and infrastructure monitoring. It’s sold as a one-stop-shop for network and system managers seeking an intelligent software agent. Dynatrace’s shares have been showing consistent growth over a long term. The stock is up a robust 133% in the past 12 months, and revenues have also been growing over that period. In the most recent report, for Q3 fiscal year 2021, the company showed $182.9 million in top-line revenue, beating the forecast by ~6% and growing 27% year-over-year. EPS came in at 6 cents, flat from Q2 and far better than the break-even reported for the year-ago quarter. Three key metrics stand out in the quarterly report, and both for the right reasons. Subscription revenue grew 33% year-over-year, to reach $170.3 million, and annual recurring revenue (ARR) – which is an important predictor of future performance – grew 35% yoy and came in at $722 million. At the same time, license revenue dropped by more than 93%, to just $300,000. Taken all together, these results point toward a strong shift toward recurring cloud customers – a common trend in the software space. Needham’s 5-star analyst Jack Andrews has been closely following Dynatrace, and he believes DT’s AI products may replace incumbent tools as customers expand to additional modules. “Embedded AIOps and automation creates a compelling value proposition… Compared to competitors in the market, DT's AI Engine is embedded within its core platform and can be levered across the portfolio to deliver answers from data. Moreover, its One Agent technology automatically discovers high-fidelity data from applications and thus can map the billions of dependencies in complex environments," Andrews said. The analyst summed up, "In our view, DT is well-positioned to serve as a single source of truth that can help users trace a line between written code and business outcomes (i.e. BizDevSecOps)." Andrews named Dynatrace as a top pick, and in line with this upbeat assessment, the analyst rates the stock a Buy along with a $66 price target. Ivestors stand to pocket ~28% gain should the analyst's thesis play out. (To watch Andrews’ track record, click here) Once again, we’re looking at a stock who strong performance has inspired unanimity from the Wall Street analysts. DT shares have 13 Buy reviews, for a Strong Buy consensus rating. The stock sells for $51.76 and its $59.69 average price target suggests ~15% upside from that level. (See DT stock analysis on TipRanks) To find good ideas for AI stocks trading at attractive valuations, visit TipRanks’ Best Stocks to Buy, a newly launched tool that unites all of TipRanks’ equity insights. Disclaimer: The opinions expressed in this article are solely those of the featured analysts. The content is intended to be used for informational purposes only. It is very important to do your own analysis before making any investment.

James Hendler is the Director of the Institute for Data Exploration and Applications and the Tetherless World Professor of Computer, Web and Cognitive Sciences at RPI. He also is acting director of the RPI-IBM Artificial Intelligence Research Collaboration and serves on the Board of the UK's charitable Web Science Trust. Hendler has authored over 400 books, technical papers and articles in the areas of Semantic Web, artificial intelligence, agent-based computing and high-performance processing. Hendler was the recipient of a 1995 Fulbright Foundation Fellowship, is a former member of the US Air Force Science Advisory Board, and is a Fellow of the AAAI, BCS, the IEEE, the AAAS and the ACM. He is also the former Chief Scientist of the Information Systems Office at the US Defense Advanced Research Projects Agency (DARPA) and was awarded a US Air Force Exceptional Civilian Service Medal in 2002.