arXiv subjects

Aditya Dave

Publications and source records attributed to Aditya Dave.

At least 19 recordsLinked to original sources

arXiv2024

Climate AI for Corporate Decarbonization Metrics Extraction

Corporate Greenhouse Gas (GHG) emission targets are important metrics in sustainable investing [12, 16]. To provide a comprehensive view of company emission objectives, we propose an approach to source these metrics from company public disclosures. Without automation, curating these metrics manually is a labor-intensive process that requires combing through lengthy corporate sustainability disclosures that often do not follow a standard format. Furthermore, the resulting dataset needs to be validated thoroughly by Subject Matter Experts (SMEs), further lengthening the time-to-market. We introduce the Climate Artificial Intelligence for Corporate Decarbonization Metrics Extraction (CAI) model and pipeline, a novel approach utilizing Large Language Models (LLMs) to extract and validate linked metrics from corporate disclosures. We demonstrate that the process improves data collection efficiency and accuracy by automating data curation, validation, and metric scoring from public corporate disclosures. We further show that our results are agnostic to the choice of LLMs. This framework can be applied broadly to information extraction from textual data.

q-fin.PM↗

arXiv2024

On Mobility Equity and the Promise of Emerging Transportation Systems

This paper introduces a mobility equity metric (MEM) for evaluating fairness and accessibility in multi-modal intelligent transportation systems. The MEM simultaneously accounts for service accessibility and transportation costs across different modes of transportation and social demographics. We provide a data-driven validation of the proposed MEM to characterize the impact of various parameters in the metric across cities in the U.S. We subsequently develop a routing framework that aims to optimize MEM within a transportation network containing both public transit and private vehicles. Within this framework, a system planner provides routing suggestions to vehicles across all modes of transportation to maximize MEM. We evaluate our approach through numerical simulations, analyzing the impact of travel demands and compliance of private vehicles. This work provides insights into designing transportation systems that are not only efficient but also equitable, ensuring fair access to essential services across diverse populations.

eess.SY↗

arXiv2024

A Mobility Equity Metric for Multi-Modal Intelligent Transportation Systems

In this paper, we introduce a metric to evaluate the equity in mobility and a routing framework to enhance the metric within multi-modal intelligent transportation systems. The mobility equity metric (MEM) simultaneously accounts for service accessibility and transportation costs to quantify the equity and fairness in a transportation network. Finally, we develop a system planner integrated with MEM that aims to distribute travel demand for the transportation network, resulting in a socially optimal mobility system. Our framework results in a transportation network that is efficient in terms of travel time, improves accessibility, and ensures equity in transportation.

eess.SY↗

arXiv2024

Safe Merging in Mixed Traffic with Confidence

In this letter, we present an approach for learning human driving behavior, without relying on specific model structures or prior distributions, in a mixed-traffic environment where connected and automated vehicles (CAVs) coexist with human-driven vehicles (HDVs). We employ conformal prediction to obtain theoretical safety guarantees and use real-world traffic data to validate our approach. Then, we design a controller that ensures effective merging of CAVs with HDVs with safety guarantees. We provide numerical simulations to illustrate the efficacy of the control approach.

eess.SY↗

arXiv2024

A Framework for Effective AI Recommendations in Cyber-Physical-Human Systems

Many cyber-physical-human systems (CPHS) involve a human decision-maker who may receive recommendations from an artificial intelligence (AI) platform while holding the ultimate responsibility of making decisions. In such CPHS applications, the human decision-maker may depart from an optimal recommended decision and instead implement a different one for various reasons. In this letter, we develop a rigorous framework to overcome this challenge. In our framework, we consider that humans may deviate from AI recommendations as they perceive and interpret the system's state in a different way than the AI platform. We establish the structural properties of optimal recommendation strategies and develop an approximate human model (AHM) used by the AI. We provide theoretical bounds on the optimality gap that arises from an AHM and illustrate the efficacy of our results in a numerical example.

eess.SY↗

arXiv2023

Stochastic Control with Distributionally Robust Constraints for Cyber-Physical Systems Vulnerable to Attacks

In this paper, we investigate the control of a cyber-physical system (CPS) while accounting for its vulnerability to external attacks. We formulate a constrained stochastic problem with a robust constraint to ensure robust operation against potential attacks. We seek to minimize the expected cost subject to a constraint limiting the worst-case expected damage an attacker can impose on the CPS. We present a dynamic programming decomposition to compute the optimal control strategy in this robust-constrained formulation and prove its recursive feasibility. We also illustrate the utility of our results by applying them to a numerical simulation.

math.OC↗

arXiv2023

A Q-learning Approach for Adherence-Aware Recommendations

In many real-world scenarios involving high-stakes and safety implications, a human decision-maker (HDM) may receive recommendations from an artificial intelligence while holding the ultimate responsibility of making decisions. In this letter, we develop an "adherence-aware Q-learning" algorithm to address this problem. The algorithm learns the "adherence level" that captures the frequency with which an HDM follows the recommended actions and derives the best recommendation policy in real time. We prove the convergence of the proposed Q-learning algorithm to the optimal value and evaluate its performance across various scenarios.

cs.LG↗

arXiv2023

Routing in Mixed Transportation Systems for Mobility Equity

This letter proposes a routing framework in mixed transportation systems for improving mobility equity. We present a strategic routing game that governs interactions between compliant and noncompliant vehicles, where noncompliant vehicles are modeled with cognitive hierarchy theory. Then, we introduce a mobility equity metric (MEM) to quantify the accessibility and fairness in the transportation network. We integrate the MEM into the routing framework to optimize it with adjustable weights for different transportation modes. The proposed approach bridges the gap between technological advancements and societal goals in mixed transportation systems to enhance efficiency and equity. We provide numerical examples and analysis of the results.

eess.SY↗

arXiv2023

Performance-Sensitive Potential Functions for Efficient Flow of Connected and Automated Vehicles

Connected and automated vehicles (CAVs) provide the most intriguing opportunity for enabling users to monitor transportation network conditions and make better decisions for improving safety and transportation efficiency. In this paper, we address the problem of effectively coordinating CAVs on lane-based roadways. Our approach utilizes potential functions to generate repulsive forces between CAVs that ensure collision avoidance. However, such potential functions can lead to unrealistic acceleration profiles and large inter-vehicle distances. The primary contribution of this work is the introduction of performance-sensitive potential functions to address these challenges. In our approach, the parameters of a potential function are determined through an optimization problem aiming to reduce both acceleration and inter-vehicle distances. To circumvent the computational implications due to the complexity of the resulting optimization problem that prevents the derivation of a real-time solution, we train a neural network model to learn the mapping of initial conditions to optimal parameters derived offline. Then, we prove sufficient criteria for the sampled-data model to ensure that the neural network output does not activate any of the state and safety constraints. Finally, we provide simulation results to demonstrate the effectiveness of the proposed approach.

eess.SY↗

arXiv2023

Connected and Automated Vehicles in Mixed-Traffic: Learning Human Driver Behavior for Effective On-Ramp Merging

Highway merging scenarios featuring mixed traffic conditions pose significant modeling and control challenges for connected and automated vehicles (CAVs) interacting with incoming on-ramp human-driven vehicles (HDVs). In this paper, we present an approach to learn an approximate information state model of CAV-HDV interactions for a CAV to maneuver safely during highway merging. In our approach, the CAV learns the behavior of an incoming HDV using approximate information states before generating a control strategy to facilitate merging. First, we validate the efficacy of this framework on real-world data by using it to predict the behavior of an HDV in mixed traffic situations extracted from the Next-Generation Simulation repository. Then, we generate simulation data for HDV-CAV interactions in a highway merging scenario using a standard inverse reinforcement learning approach. Without assuming a prior knowledge of the generating model, we show that our approximate information state model learns to predict the future trajectory of the HDV using only observations. Subsequently, we generate safe control policies for a CAV while merging with HDVs, demonstrating a spectrum of driving behaviors, from aggressive to conservative. We demonstrate the effectiveness of the proposed approach by performing numerical simulations.

cs.LG↗

arXiv2023

A Study of an Atomic Mobility Game With Uncertainty Under Prospect Theory

In this paper, we present a study of a mobility game with uncertainty in the decision-making of travelers and incorporate prospect theory to model travel behavior. We formulate a mobility game that models how travelers distribute their traffic flows in a transportation network with splittable traffic, utilizing the Bureau of Public Roads function to establish the relationship between traffic flow and travel time cost. Given the inherent non-linearities and complexity introduced by the uncertainties, we propose a smooth approximation function to estimate the prospect-theoretic cost functions. As part of our analysis, we characterize the best-fit parameters and derive an upper bound for the error. We then show the existence of an equilibrium and its its best-possible approximation.

eess.SY↗

arXiv2023

Worst-Case Control and Learning Using Partial Observations Over an Infinite Time-Horizon

Safety-critical cyber-physical systems require control strategies whose worst-case performance is robust against adversarial disturbances and modeling uncertainties. In this paper, we present a framework for approximate control and learning in partially observed systems to minimize the worst-case discounted cost over an infinite time horizon. We model disturbances to the system as finite-valued uncertain variables with unknown probability distributions. For problems with known system dynamics, we construct a dynamic programming (DP) decomposition to compute the optimal control strategy. Our first contribution is to define information states that improve the computational tractability of this DP without loss of optimality. Then, we describe a simplification for a class of problems where the incurred cost is observable at each time instance. Our second contribution is defining an approximate information state that can be constructed or learned directly from observed data for problems with observable costs. We derive bounds on the performance loss of the resulting approximate control strategy and illustrate the effectiveness of our approach in partially observed decision-making problems with a numerical example.

math.OC↗

arXiv2023

Approximate Information States for Worst-Case Control and Learning in Uncertain Systems

In this paper, we investigate discrete-time decision-making problems in uncertain systems with partially observed states. We consider a non-stochastic model, where uncontrolled disturbances acting on the system take values in bounded sets with unknown distributions. We present a general framework for decision-making in such problems by using the notion of the information state and approximate information state, and introduce conditions to identify an uncertain variable that can be used to compute an optimal strategy through a dynamic program (DP). Next, we relax these conditions and define approximate information states that can be learned from output data without knowledge of system dynamics. We use approximate information states to formulate a DP that yields a strategy with a bounded performance loss. Finally, we illustrate the application of our results in control and reinforcement learning using numerical examples.

eess.SY↗

arXiv2022

On Robust Control of Partially Observed Uncertain Systems with Additive Costs

In this paper, we consider the problem of optimizing the worst-case behavior of a partially observed system. All uncontrolled disturbances are modeled as finite-valued uncertain variables. Using the theory of cost distributions, we present a dynamic programming (DP) approach to compute a control strategy that minimizes the maximum possible total cost over a given time horizon. To improve the computational efficiency of the optimal DP, we introduce a general definition for information states and show that many information states constructed in previous research efforts are special cases of ours. Additionally, we define approximate information states and an approximate DP that can further improve computational tractability by conceding a bounded performance loss. We illustrate the utility of these results using a numerical example.

math.OC↗

arXiv2022

Approximate Information States for Worst-case Control of Uncertain Systems

In this paper, we investigate a worst-case-scenario control problem with a partially observed state. We consider a non-stochastic formulation, where noises and disturbances in our dynamics are uncertain variables which take values in finite sets. In such problems, the optimal control strategy can be derived using a dynamic program (DP) with respect to the memory. The computational complexity of this DP can be improved using a conditional range of the state instead of the memory. We present a more general definition of an information state which is sufficient to construct a DP without loss of optimality, and show that the conditional range is an example of an information state. Next, we extend this notion to define an approximate information state and an approximate DP. We also bound the maximum loss of optimality when using an approximate DP to derive the control strategy. Finally, we illustrate our results in a numerical example.

math.OC↗

arXiv2021

Decentralized Control of Two Agents with Nested Accessible Information

In this paper, we investigate a decentralized stochastic control problem with two agents, where a part of the memory of the second agent is also available to the first agent at each instance of time. We derive a structural form for optimal control strategies which allows us to restrict their domain to a set which does not grow in size with time. We also present a dynamic programming (DP) decomposition which can utilize our results to derive optimal strategies for arbitrarily long time horizons. Since obtaining optimal control strategies by solving this DP decomposition is computationally intensive, we present potential resolutions in the form of simplified strategies by imposing additional conditions on our model, and an approximation technique which can be used to implement our results with a bounded loss of optimality.

math.OC↗

arXiv2021

On Decentralized Minimax Control with Nested Subsystems

In this paper, we investigate a decentralized control problem with nested subsystems, which is a general model for one-directional communication amongst many subsystems. The noises in our dynamics are modelled as uncertain variables which take values in finite sets. The objective is to minimize a worst-case shared cost. We demonstrate how the prescription approach can simplify the information structure and derive a structural form for optimal control strategies. The structural form allows us to restrict attention to control strategies whose domains do not grow in size with time, and thus, this form can be utilized in systems with long time horizons. Finally, we present a dynamic program to derive the optimal control strategies and validate our results with a numerical example.

math.OC↗

arXiv2021

A Dynamic Program for a Team of Two Agents with Nested Information

In this paper, we investigate a sequential dynamic team problem consisting of two agents with a nested information structure. We use a combination of the person-by-person and prescription approach to derive structural results for optimal control strategies for the team. We then use these structural results to present a dynamic programming (DP) decomposition to derive the optimal control strategies for a finite time horizon. We show that our DP utilizes the nested information structure to simplify the computation of the optimal control laws for the team at the final time step.

math.OC↗