Insights generated by advanced analytics in smart grids are critical for delivering safe and reliable power and providing value-added services in a cost-effective manner. However, electricity consumption data collected from advanced metering infrastructure (AMI) often contain missing values due to measurement errors, communication failures, device outages, or other unknown causes. Thus, missing data imputation is critical for ensuring the reliability of data-driven analysis. The existing literature has applied various imputation methods for electricity consumption data, but systematic comparison across different method families, missingness rates and contiguous gap sizes remain limited. In this work, we conducted a comparative analysis of eleven widely used imputation methods, including persistence-based, fixed statistical methods, time-series based, and machine learning based approaches, using high-resolution electricity consumption data from Slovenian households. The methods are evaluated using normalized root mean square error (NRMSE) as the key performance metric. Statistical significance testing is further used to determine whether the observed performance differences between the leading methods are consistent across households and experimental settings. The results show that the performance of imputation methods is strongly influenced by both missingness rate and gap size. While linear interpolation performs well for very short gaps, methods capturing historical and contextual consumption patterns perform better for medium and long gaps. Among the eleven evaluated methods, XGBoost achieves the best overall performance, with the lowest average NRMSE. Pairwise Wilcoxon signed-rank tests further confirm that XGBoost significantly outperforms the leading competing methods in most evaluated configurations. The consumer level analysis also shows that the most accurate method on average is not always the most consistent across households. Thus, researchers and practitioners should consider gap size, temporal structure, and consumer-level variability, in addition to the percentage of missing values, when selecting imputation methods for smart meter data.
This paper comprehensively surveys research trends in imitation learning (IL) for contact-rich robotic tasks. Contact-rich tasks, which require complex physical interactions with the environment, represent a central challenge in robotics due to their nonlinear dynamics and sensitivity to small positional deviations. The paper examines demonstration collection methodologies, including teaching methods and sensory modalities crucial for capturing subtle interaction dynamics. We then analyze IL approaches, highlighting their applications to contact-rich manipulation. Recent advances in multimodal learning and foundation models have significantly enhanced performance in complex contact tasks across industrial, household, and healthcare domains. Through systematic organization of current research and identification of challenges, this survey provides a foundation for future advancements in contact-rich robotic manipulation.
Recently there has been an increased interest in possible tests of locality via Bell’s inequalities, as well as separately tests of entanglement at colliders, in particular at the LHC. These have involved various physical processes, such as tt , or τ+τ− production, or the decay of a Higgs boson to two vector bosons H → VV∗. We argue that none of these proposals constitute a test of locality via Bell’s inequality or promise unconditional observational evidence of entanglement. In all cases what is measured are the momenta of the final state particles. Using the construction proposed by Kasday (1971) in a different context, and adapted to collider scenarios by Abel, Dittmar, and Dreiner (1992), it is straightforward to construct a local hidden variable theory (LHVT) which exactly reproduces the data. This construction is only possible as the final state momenta all commute. We show that this LHVT satisfies Bell’s inequality or the related CHSH inequality as appropriate for all the proposed LHC collider tests in the literature. Thus a test of locality via Bell’s inequality is not possible at colliders. The LHVT is also by construction local, i.e. all correlations are separable. Thus an unconditional proof of entanglement is also inherently not possible at colliders. It can only be shown that the entanglement within the Standard Model consistently describes the data.
Artificial Intelligence (AI) is being incorporated in several optimization, scheduling, orchestration as well as in native communication network functions. This paradigm shift results in increased energy consumption, however, quantifying the end-to-end energy consumption of adding intelligence to communication systems remains an open challenge since conventional energy consumption metrics focus on either communication, computation infrastructure, or model development. To address this, we propose a new metric, the Energy Cost of AI Lifecycle (eCAL) of an AI model in a system. eCAL captures the energy consumption throughout the development, deployment and utilization of an AI-model providing intelligence in a communication network by (i) analyzing the complexity of data collection and manipulation in individual components and (ii) deriving overall and per-bit energy consumption. We show that as a trained AI model is used more frequently for inference, its energy cost per inference decreases, since the fixed training energy is amortized over a growing number of inferences. For a simple case study we show that eCAL for 100 inferences is 2.73 times higher than for 1000 inferences. Additionally, we have developed a modular and extendable open-source simulation tool to enable researchers, practitioners, and engineers to calculate the end-to-end energy cost with various configurations and across various systems, ensuring adaptability to diverse use cases.
Background:Reliable nutrient profiling and semantic interoperability are essential for scalable dietary assessment, food labeling (e.g., traffic-light schemes), and FAIR integration of food composition and consumption data. However, general-purpose large language models (LLMs) are not systematically exposed to structured recipe-nutrition mappings and food ontologies, limiting their accuracy and trustworthiness in food and nutrition tasks. Scope and approach:We review recent LLM advances in life sciences and healthcare and analyze the gap in food and nutrition applications. To address this gap, we introduce FoodyLLM, a domain-specialized LLM fine-tuned on 225k task-aligned QA pairs for (i) recipe nutrient estimation, (ii) traffic-light classification, and (iii) ontology-based entity linking to support FAIR food data interoperability. We benchmark FoodyLLM against strong general-purpose baselines (e.g., Llama 3 8B, Gemini 2.0) under zero-/few-shot prompting across five evaluation folds. Key findings:Across all tasks, FoodyLLM substantially outperforms general-purpose LLMs for nutrient estimation across all macronutrients (fat, protein, salt, saturates, sugar), accuracy increases from 0.43 to 0.63 to 0.91-0.97; for traffic-light classification across all nutrients and color categories, macro F1 improves from 0.46 to 0.80 to 0.86-0.97; and for ontology-based food entity linking across FoodOn, SNOMED-CT, and Hansard, macro F1 increases from 0.33 to 0.44 (best general-purpose baseline) to 0.93-0.98 on artificial NEL data, and from 0.24 to 0.51 to 0.67-0.84 on real corpora (CafeteriaSA and CafeteriaFCD). Overall, our results demonstrate the practical value of domain-specialized LLMs in food and nutrition research. They enable automated dietary assessment, large-scale nutritional monitoring, and FAIR data integration, while opening new pathways toward sustainable and personalized nutrition.
