Costas J. Spanos
YOU?
Author Swipe
View article: StyleBench: Evaluating thinking styles in Large Language Models
StyleBench: Evaluating thinking styles in Large Language Models Open
The effectiveness of Large Language Models (LLMs) is heavily influenced by the reasoning strategies, or styles of thought, employed in their prompts. However, the interplay between these reasoning styles, model architecture, and task type …
View article: Enhancing LLM Reasoning for Time Series Classification by Tailored Thinking and Fused Decision
Enhancing LLM Reasoning for Time Series Classification by Tailored Thinking and Fused Decision Open
The reasoning capabilities of large language models (LLMs) have significantly advanced their performance by enabling in-depth understanding of diverse tasks. With growing interest in applying LLMs to the time series domain, this has proven…
View article: Safe Continual Domain Adaptation after Sim2Real Transfer of Reinforcement Learning Policies in Robotics
Safe Continual Domain Adaptation after Sim2Real Transfer of Reinforcement Learning Policies in Robotics Open
Domain randomization has emerged as a fundamental technique in reinforcement learning (RL) to facilitate the transfer of policies from simulation to real-world robotic applications. Many existing domain randomization approaches have been p…
View article: Robust Gymnasium: A Unified Modular Benchmark for Robust Reinforcement Learning
Robust Gymnasium: A Unified Modular Benchmark for Robust Reinforcement Learning Open
Driven by inherent uncertainty and the sim-to-real gap, robust reinforcement learning (RL) seeks to improve resilience against the complexity and variability in agent-environment sequential interactions. Despite the existence of a large nu…
View article: Don't Trade Off Safety: Diffusion Regularization for Constrained Offline RL
Don't Trade Off Safety: Diffusion Regularization for Constrained Offline RL Open
Constrained reinforcement learning (RL) seeks high-performance policies under safety constraints. We focus on an offline setting where the agent has only a fixed dataset -- common in realistic tasks to prevent unsafe exploration. To addres…
View article: Enhancing Efficiency of Safe Reinforcement Learning via Sample Manipulation
Enhancing Efficiency of Safe Reinforcement Learning via Sample Manipulation Open
Safe reinforcement learning (RL) is crucial for deploying RL agents in real-world applications, as it aims to maximize long-term rewards while satisfying safety constraints. However, safe RL often suffers from sample inefficiency, requirin…
View article: Active Reinforcement Learning for Robust Building Control
Active Reinforcement Learning for Robust Building Control Open
Reinforcement learning (RL) is a powerful tool for optimal control that has found great success in Atari games, the game of Go, robotic control, and building optimization. RL is also very brittle; agents often overfit to their training env…
View article: Machine Learning for Smart and Energy-Efficient Buildings
Machine Learning for Smart and Energy-Efficient Buildings Open
Energy consumption in buildings, both residential and commercial, accounts for approximately 40% of all energy usage in the United States, and similar numbers are being reported from countries around the world. This significant amount of e…
View article: Active Reinforcement Learning for Robust Building Control
Active Reinforcement Learning for Robust Building Control Open
Reinforcement learning (RL) is a powerful tool for optimal control that has found great success in Atari games, the game of Go, robotic control, and building optimization. RL is also very brittle; agents often overfit to their training env…
View article: Personalized Federated Hypernetworks for Multi-Task Reinforcement Learning in Microgrid Energy Demand Response
Personalized Federated Hypernetworks for Multi-Task Reinforcement Learning in Microgrid Energy Demand Response Open
As sensors pervade the built environment, they have fueled the advance of data-driven models that promise greater efficiency for microgrid management. However, this has raised concerns over data privacy and data ownership. The paradigm of …
View article: Energy savings and thermal comfort in a zero energy office building with fans in Singapore
Energy savings and thermal comfort in a zero energy office building with fans in Singapore Open
Elevated air movement produced by fans can offset air-conditioning energy requirements by allowing temperature setpoints to be raised without compromising thermal comfort. These advantages are even greater in hot and humid climates that in…
View article: Time Varying Marginal Emissions Intensity of Energy Consumption: Implications for Flexible Loads
Time Varying Marginal Emissions Intensity of Energy Consumption: Implications for Flexible Loads Open
Climate-conscious electricity consumers can modify their energy consumption patterns by shifting or shedding load in order to reduce carbon emissions. The impact of modified consumption on emissions is through the marginal emissions intens…
View article: Toward Platform-based Building Design
Toward Platform-based Building Design Open
The electronic design industry has undergone a significant transformation, transitioning from traditional hand-drawn designs to modern automated design processes. While Computer-Aided Design (CAD) tools emerged alongside the electronic ind…
View article: From Electronic Design Automation to Building Design Automation: Challenges and Opportunities
From Electronic Design Automation to Building Design Automation: Challenges and Opportunities Open
Design automation, which involves the use of software tools and technologies to streamline the design process, has been widely adopted in the electronics industry, resulting in significant advancements in product development and manufactur…
View article: Are There Differences between the Stress Responses of Philippine Men and Women to the COVID-19 Pandemic?
Are There Differences between the Stress Responses of Philippine Men and Women to the COVID-19 Pandemic? Open
The SARS-CoV-2 pandemic has had a deleterious impact on human health since its beginning in 2019. The purpose of this study was to examine the psychosocial impact of the COVID-19 pandemic in the Philippines and determine if there were diff…
View article: Energy Savings and Thermal Comfort in a Zero Energy Office Building with Fans in Singapore
Energy Savings and Thermal Comfort in a Zero Energy Office Building with Fans in Singapore Open
View article: Machine Learning for Smart and Energy-Efficient Buildings
Machine Learning for Smart and Energy-Efficient Buildings Open
Energy consumption in buildings, both residential and commercial, accounts for approximately 40% of all energy usage in the U.S., and similar numbers are being reported from countries around the world. This significant amount of energy is …
View article: Improved dequantization and normalization methods for tabular data pre-processing in smart buildings
Improved dequantization and normalization methods for tabular data pre-processing in smart buildings Open
Ubiquitous deployment of IoT sensors marks a defining characteristic of smart buildings, for they constitute the source of data on building operation, diagnosis, and maintenance. For machine learning applications in buildings, often the se…
View article: Synthetic personal thermal comfort data generation
Synthetic personal thermal comfort data generation Open
Personal thermal comfort models aim to predict an individual's thermal comfort response, instead of the average response of a large group. Recently, machine learning algorithms have proven to be having enormous potential as a candidate for…
View article: Optimizing participation of buildings and aggregations in incentive-based demand response programs
Optimizing participation of buildings and aggregations in incentive-based demand response programs Open
Demand response (DR) programs are used to modulate electricity load in situations when procuring additional electricity to meet peak demand is expensive, and comes from polluting sources like natural gas peaker plants. At these times, it i…
View article: Adversarial poisoning attacks on reinforcement learning-driven energy pricing
Adversarial poisoning attacks on reinforcement learning-driven energy pricing Open
Complex controls are increasingly common in power systems. Reinforcement learning (RL) has emerged as a strong candidate for implementing various controllers. One common use of RL in this context is for prosumer pricing aggregations, where…
View article: Personalized Federated Hypernetworks for Privacy Preservation in Multi-Task Reinforcement Learning
Personalized Federated Hypernetworks for Privacy Preservation in Multi-Task Reinforcement Learning Open
Multi-Agent Reinforcement Learning currently focuses on implementations where all data and training can be centralized to one machine. But what if local agents are split across multiple tasks, and need to keep data private between each? We…
View article: Deep reinforcement learning with planning guardrails for building energy demand response
Deep reinforcement learning with planning guardrails for building energy demand response Open
Building energy demand response is projected to be important in decarbonizing energy use. A demand response program that communicates “artificial” hourly price signals to workers as part of a social game has the potential to elicit energy …
View article: A Survey of ADMM Variants for Distributed Optimization: Problems, Algorithms and Features
A Survey of ADMM Variants for Distributed Optimization: Problems, Algorithms and Features Open
By coordinating terminal smart devices or microprocessors to engage in cooperative computation to achieve systemlevel targets, distributed optimization is incrementally favored by both engineering and computer science. The well-known alter…
View article: Conditional Synthetic Data Generation for Robust Machine Learning Applications with Limited Pandemic Data
Conditional Synthetic Data Generation for Robust Machine Learning Applications with Limited Pandemic Data Open
Background: At the onset of a pandemic, such as COVID-19, data with proper labeling/attributes corresponding to the new disease might be unavailable or sparse. Machine Learning (ML) models trained with the available data, which is limited …
View article: Optimal Network Charge for Peer-to-Peer Energy Trading: A Grid Perspective
Optimal Network Charge for Peer-to-Peer Energy Trading: A Grid Perspective Open
Peer-to-peer (P2P) energy trading is a promising market scheme to accommodate the increasing distributed energy resources (DERs). However, how P2P to be integrated into the existing power systems remains to be investigated. In this paper, …
View article: Time series-based deep learning model for personal thermal comfort prediction
Time series-based deep learning model for personal thermal comfort prediction Open
Personal thermal comfort models are crucial for the future of human-in-the-loop HVAC control in energy-efficient buildings. Individual comfort models, compared to average population responses, can provide the personalization required for s…
View article: Proximal ADMM for Nonconvex and Nonsmooth Optimization
Proximal ADMM for Nonconvex and Nonsmooth Optimization Open
By enabling the nodes or agents to solve small-sized subproblems to achieve coordination, distributed algorithms are favored by many networked systems for efficient and scalable computation. While for convex problems, substantial distribut…
View article: Optimal Network Charge for Peer-to-Peer Energy Trading: A Grid Perspective
Optimal Network Charge for Peer-to-Peer Energy Trading: A Grid Perspective Open
Peer-to-peer (P2P) energy trading is a promising market scheme to accommodate the increasing distributed energy resources (DERs). However, how P2P to be integrated into the existing power systems remains to be investigated. In this paper, …
View article: smartSDH: An Experimental Study of Mechanism-Based Building Control
smartSDH: An Experimental Study of Mechanism-Based Building Control Open
As Internet of Things technologies are increasingly being deployed, situations frequently arise where multiple stakeholders must reconcile preferences to control a shared resource. We perform a five-month long experiment dubbed “smartSDH” …