Cloud Computing and Data Science

Taming and Controlling Performance and Energy Trade-Offs Automatically in Network Applications

Han Dong — 2026-03-11

In this paper, we demonstrate that a server running a single latency-sensitive application can be treated as a black box to reduce energy consumption while meeting a Service-Level Agreement (SLA) target. We find that it is possible to identify “sweet spot” settings for packet batching and processing rate control. These settings represent optimal trade-offs between the software stack and hardware. Specifically, they account for both the arrival rate and the composition of requests being served. By testing a few combinations of these settings on the live system, a proof-of concept controller can dynamically find settings that reduce energy consumption while meeting a desired tail latency for the request rate. Our work demonstrates three key findings. First, without software changes, energy savings of up to 60% are achievable across diverse hardware systems by controlling batching and processing rates. Second, specialized research Operating Systems (OSes) can leverage this to achieve a further 40% energy savings over general-purpose OSes. Finally, we show that a controller that is agnostic to the application, system, and hardware, can find energy efficient settings for different request rates while meeting performance objectives.

TinyML-Based Federated Learning: A Novel Framework for Privacy-Preserving Smart Healthcare Applications

Manas Kumar Yogi — 2026-04-02

This paper presents an optimized integration framework combining Tiny Machine Learning (TinyML) and Federated Learning (FL) for privacy-preserving smart healthcare applications. While building upon established techniques, our contribution lies in their synergistic adaptation and optimization for resource-constrained healthcare Internet of Things (IoT) environments. We implement Adaptive Noise Injection (ANI) with data-sensitive tuning and Authenticated Homomorphic Encryption (AHE) using the Cheon-Kim-Kim-Song (CKKS) scheme to create a multi-layered privacy shield. Experimental validation using synthetic Electronic Health Record (EHR) data (derived from real Indonesian hospital patterns) demonstrates an effective privacy-utility balance, achieving 89% classification accuracy with differential privacy (ε = 1.0, σ = 0.01). The framework maintains inference latency under 60 ms with only 5% estimated daily battery consumption on typical wearable hardware.

A K-Means, Ward, and DBSCAN Repeatability Study

Anthony Bertrand — 2026-04-02

Reproducibility is essential in machine learning because it ensures that a model or experiment yields the same scientific conclusion. For specific algorithms, repeatability with bitwise identical results is also a key for scientific integrity because it allows debugging. We decomposed several very popular clustering algorithms: K-Means, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), and Ward into their fundamental steps, and we identify the conditions required to achieve repeatability at each stage. We use an implementation example with the Python library scikit-learn to examine the repeatable aspects of each method. Our results reveal non-repeatable behavior with K-Means when the number of OpenMP threads exceeds two. This work aims to raise awareness of this issue among both users and developers, encouraging further investigation and potential fixes.

The Personalization Paradox: Semantic Loss Vs. Reasoning Gains in Agentic AI Q & A

Satyajit Movidi — 2026-05-18

This study examines how personalization in agentic retrieval-augmented Artificial Intelligence (AI) systems influences the quality of answers delivered in institutional knowledge access settings such as academic advising. Prior advising and knowledge-access systems typically assume personalization is universally beneficial, yet little empirical evidence evaluates how it alters information quality. This paper addresses this gap by analyzing personalization as an independent factor within a Retrieval-Augmented Generation Large Language Model (RAG LLM) used for student advising. The study evaluates ten system configurations across personalized and non-personalized conditions using twelve authentic advising questions intentionally designed for lexical strictness. Performance was assessed using lexical metrics (Bilingual Evaluation Understudy (BLEU), Recall-Oriented Understudy for Gisting Evaluation (ROUGE)-L), semantic similarity measures (Metric for Evaluation of Translation with Explicit ORdering (METEOR), BERTScore), and reasoning/grounding metrics from the Retrieval Augmented Generation Assessment (RAGAs) framework. A Linear Mixed-effects Model (LMM) was used to quantify main effects and interactions. Personalization in agentic AI does not yield uniform gains; instead, it creates a critical trade-off where factors that significantly improve reasoning quality also incur a statistically significant penalty on semantic similarity. Specifically, personalized configurations produced a statistically significant decrease in BERTScore (0.841 vs. 0.848, p < 0.0001) alongside a simultaneous and significant improvement in METEOR (0.361 vs. 0.251, Δ = + 0.110, p < 0.0001), together demonstrating the metric-dependent nature of the trade-off. Grounding and reasoning metrics simultaneously improved, with Faithfulness rising from 0.655 to 0.711 (p = 0.0135), further supporting that personalization enhances answer quality even as it is penalized by semantic similarity metrics. Personalization decreases semantic similarity scores, not due to quality loss but because generic semantic metrics penalize beneficial user-specific deviations. The configuration that applied personalization redundantly across all three stages: setting the AI’s role, guiding document retrieval, and conditioning the final response generation, achieved the best overall results, confirming that fully integrated user-specific adaptation yields the most effective balance between reasoning gains and semantic penalties.

Advancing Explainable AI for Clinical Decision Support: A Multimodal Evaluation Framework

Hashim Ali — 2026-04-23

The increasing integration of Artificial Intelligence (AI) into Clinical Decision Support Systems (CDSS) is constrained by the limited transparency of model predictions, which undermines clinician trust and slows adoption in safety-critical settings. To address this barrier, we propose a unified multimodal evaluation framework for eXplainable AI (XAI) and empirically assess the behavior of the explanations in four clinically relevant modalities: chest radiography for pathology detection, Electroencephalography (EEG)-based epilepsy decision support using High-Frequency Oscillation (HFO) evidence, multimodal emotion recognition for psychological decision support and prediction of Alzheimer's disease based on Electronic Health Records (EHR). The framework evaluates widely used explanation mechanisms, including Gradient-weighted Class Activation Mapping (Grad-CAM), Integrated Gradients (IG), SHapley Additive exPlanations (SHAP)-style feature attribution, and attention-based interpretation, using modality-appropriate criteria that emphasize reliability, robustness, and clinical plausibility rather than visualization quality alone. The results show that Grad-CAM provides stable region-level localization in chest X-ray prediction. In contrast, EEG-based epilepsy decision support requires interpretability grounded in domain-specific biomarkers and time-frequency structure rather than generic saliency. In multimodal emotion recognition, fusion improves performance, but the contribution of each modality varies by emotional state, highlighting the need for interpretable fusion analyses. For Alzheimer's prediction, a tuned CatBoost model achieves strong discrimination, and feature-level analyses identify clinically significant drivers, including Mini-Mental State Examination (MMSE)-related measures. Cross-modal synthesis demonstrates that explanation effectiveness is inherently task- and modality-dependent and that explanation instability and susceptibility to spurious cues remain recurring risks across settings, consistent with concerns about the faithfulness of saliency explanations. In general, the proposed framework supports the practical implementation of XAI in healthcare by providing modality-aligned guidance for selecting and validating explanations within clinical workflows.

From Adoption to Execution: Challenges and Frameworks for Cloud ERP Implementation-A Systematic Literature Review

Caroline Hornegger — 2026-03-19

An increasing number of companies are migrating their Enterprise Resource Planning (ERP) systems to the cloud-an area that remains relatively underexplored, as traditional ERP systems were typically deployed on premises. While many organizations already operate other systems in the cloud, ERP systems are particularly critical because they integrate core business processes and manage daily operations, making cloud migration a high-risk transformation that must be carefully planned and executed to ensure business continuity. This study aims to identify the key challenges and implementation frameworks associated with Cloud ERP migration through a Systematic Literature Review (SLR). Studies published between 2015 and 2025 were retrieved from five academic databases, screened using predefined inclusion and quality criteria, and synthesised using Excel and Orange Data Mining software, resulting in a final sample of 58 studies. The results identify 26 distinct challenges-such as data migration, security and privacy concerns, vendor dependence, and resistance to change-and 12 classes of frameworks intended to mitigate these barriers across different organizational and contextual settings. In contrast to prior reviews that primarily catalogue adoption drivers or isolated challenges, this study contributes a structured synthesis that explicitly maps implementation challenge categories to classes of implementation frameworks across different organisational contexts. This integrative perspective reveals systematic coverage gaps in existing frameworks and provides decision-oriented guidance for selecting implementation approaches in practice.