Explainable AI: Bridging the Gap between Machine Learning Models and Human Understanding

Journal of Informatics Education and Research, Journal Year: 2024, Volume and Issue: unknown

Published: Jan. 1, 2024

Explainable AI (XAI) is one of the key game-changing features in machine learning models, which contribute to making them more transparent, regulated and usable different applications. In (the) investigation this paper, we consider four rows explanation methods—LIME, SHAP, Anchor, Decision Tree-based Explanation—in disentangling decision-making process black box models within fields. our experiments, use datasets that cover domains, for example, health, finance image classification, compare accuracy, fidelity, coverage, precision human satisfaction each method. Our work shows rule trees approach called (Decision explanation) mostly superior comparison other non-model-specific methods performing higher coverage regardless classifier. addition this, respondents who answered qualitative evaluation indicated they were very content with decision tree-based explanations these types are easy understandable. Furthermore, most famous sorts clarifications instinctive significant. The over discoveries stretch on utilize interpretable strategies facilitating hole between understanding thus advancing straightforwardness responsibility AI-driven decision-making.

Language: Английский

---

The research landscape of industry 5.0: a scientific mapping based on bibliometric and topic modeling techniques

Flexible Services and Manufacturing Journal, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 30, 2024

Abstract Industry 5.0 (I5.0) marks a transformative shift toward integrating advanced technologies with human-centric design to foster innovation, resilient manufacturing, and sustainability. This study aims examine the evolution collaborative dynamics of I5.0 research through bibliometric analysis 942 journal articles from Scopus database. Our findings reveal significant increase in research, particularly post-2020, yet highlight fragmented collaboration networks noticeable gap between institutions developed developing countries. Key thematic areas identified include human-robot collaboration, data management security, AI-driven sustainable practices. These insights suggest that more integrated approach is essential for advancing I5.0, calling strengthened global collaborations balanced emphasis on both technological elements fully realize its potential driving industrial provides first comprehensive offering valuable researchers practitioners.

Language: Английский

---

Industry 5.0 as seen through its academic literature: an investigation using co-word analysis

Discover Sustainability, Journal Year: 2025, Volume and Issue: 6(1)

Published: April 19, 2025

Language: Английский

---

Anomaly Detection in Industrial Processes: Supervised vs. Unsupervised Learning and the Role of Explainability

Open Research Europe, Journal Year: 2025, Volume and Issue: 5, P. 8 - 8

Published: Jan. 14, 2025

Background Anomaly detection is vital in industrial settings for identifying abnormal behaviors that suggest faults or malfunctions. Artificial intelligence (AI) offers significant potential to assist humans addressing these challenges. Methods This study compares the performance of supervised and unsupervised machine learning (ML) techniques anomaly detection. Additionally, model-specific explainability methods were employed interpret outputs. A novel approach, MLW-XAttentIon, based on causal reasoning attention networks, was proposed visualize inference process transformer models. Results Experimental results revealed models perform well without requiring labeled data, offering promise. In contrast, demonstrated greater robustness reliability. Conclusions Unsupervised ML present a feasible, resource-efficient option detection, while remain more reliable critical applications. The MLW-XAttentIon approach enhances interpretability transformer-based models, contributing trust transparency AI-driven systems.

Language: Английский

---

From Sensors to Digital Twins toward an Iterative Approach for Existing Manufacturing Systems

Sensors, Journal Year: 2024, Volume and Issue: 24(5), P. 1434 - 1434

Published: Feb. 23, 2024

Digital twin technology is a highly valued asset in the manufacturing sector, with its unique capability to bridge gap between physical and virtual parts. The impact of rapid increase this based on collection real-world data, standardization, widespread deployment an existing system. This encompasses sensor values, PLC internal states, IoT, as well how means linking these data their digital counterparts. It challenging implement twins large scale due heterogeneity protocols structuring subsystems. To facilitate integration into architectures, we propose paper framework that enables scalable from sensors services iterative manner.

Language: Английский

---

Analysis of Azure Zero Trust Architecture Implementation for Mid-size Organizations

Published: July 18, 2024

The Zero Trust Architecture (ZTA) security system follows the "never trust, always verify" principle. process constantly verifies users and devices trying to access resources. This paper describes how Microsoft Azure uses ZTA enforce strict identity verification rules across cloud environment improve security. Implementation is time-consuming difficult. Azure's extensive services customizations require careful design implementation. administrators struggle navigate change configurations due its complex user interface (UI). Each ecosystem component must meet criteria. ZTA's comprehensive policy definitions, multi-factor pass-wordless authentication, other advanced features are tested in a mid-size business scenario. These configuration changes solid grasp of architecture Trust. Implementing an reduces vulnerabilities restricts authorized users' critical expensive, requires significant commitment changing IT procedures, can be confusing because same available multiple places. has shown great potential for both hybrid cloud-native companies.

Language: Английский

---

Harmonising humans and technology: Exploring the dynamics of cognitive production, artificial intelligence and social communication in cybernetic systems

Open Research Europe, Journal Year: 2024, Volume and Issue: 4, P. 201 - 201

Published: Sept. 9, 2024

Agile cognitive production systems mark a manufacturing paradigm shift, propelled by the demand for accelerated product development and adoption of digitalised across extensive supply networks. Cognitive emphasises role technology automation in learning adaptation process. These independently analyse data, make real-time adjustments optimise processes, sometimes minimising need human intervention. Based on conceptual framework that draws diversity living cybernetics provides solid theoretical background. It explores intricate connections between cognition, self-organising challenges arising from autonomy such systems. The concept "cognition" "agile systems" moves away conventional understanding purely technical processes towards thought processes. This departure fosters dynamic exchange where individual thoughts resonate social communication. Addressing artificial intelligence (AI), article examining computers science standpoint, exploring relationship mental systems, capturing faculties as utterance, understanding. integration AI into computer-mediated communication leads to question how AI-equipped intersect with societal notions. inherent intransparency AI, often viewed black box, prompts queries about potential black-box nature an autonomously controlled factory or chain. In this hypothetical scenario, idea chain network is challenged, emphasising importance involvement. Research human-centric explainable human-in-the-loop. orientation goes beyond dimensions incorporates considerations, which holistic current research. essence, research field comprehensive exploration complex interplay evolving landscape modern systems.

Language: Английский

---

Industry 4.0/IIoT Platforms for manufacturing systems - A systematic review contrasting the scientific and the industrial side

Information and Software Technology, Journal Year: 2024, Volume and Issue: unknown, P. 107650 - 107650

Published: Dec. 1, 2024

Language: Английский

---

Analysis of Azure Zero Trust Architecture Implementation for Mid-Size Organizations

Journal of Cybersecurity and Privacy, Journal Year: 2024, Volume and Issue: 5(1), P. 2 - 2

Published: Dec. 30, 2024

The Zero Trust Architecture (ZTA) security system follows the “never trust, always verify” principle. process constantly verifies users and devices trying to access resources. This paper describes how Microsoft Azure uses ZTA enforce strict identity verification rules across cloud environment improve security. Implementation takes time effort. Azure’s extensive services customizations require careful design implementation. administrators need help navigating changing configurations due its complex user interface (UI). Each ecosystem component must meet criteria. ZTAs comprehensive policy definitions, multi-factor passwordless authentication, other advanced features are tested in a mid-size business scenario. document delineates several principal findings concerning execution of within mid-sized enterprises. significantly improves by reducing attack surfaces via ongoing verification, stringent controls, micro-segmentation. Nonetheless, is resource-demanding intricate, necessitating considerable expertise meticulous planning. A notable disparity exists between theoretical frameworks their practical implementation, characterized disjointed management interfaces fatigue resulting from incessant authentication requests. case studies indicate that although enhances organizational mitigates risks, it may disrupt operations adversely affect experience, particularly hybrid fully cloud-based settings. study underscores necessity for customized equilibrium usability ensure effective

Language: Английский

---