Understanding Physics-Informed Neural Networks: Techniques, Applications, Trends, and Challenges

AI, Journal Year: 2024, Volume and Issue: 5(3), P. 1534 - 1557

Published: Aug. 29, 2024

Physics-informed neural networks (PINNs) represent a significant advancement at the intersection of machine learning and physical sciences, offering powerful framework for solving complex problems governed by laws. This survey provides comprehensive review current state research on PINNs, highlighting their unique methodologies, applications, challenges, future directions. We begin introducing fundamental concepts underlying motivation integrating physics-based constraints. then explore various PINN architectures techniques incorporating laws into network training, including approaches to partial differential equations (PDEs) ordinary (ODEs). Additionally, we discuss primary challenges faced in developing applying such as computational complexity, data scarcity, integration Finally, identify promising Overall, this seeks provide foundational understanding PINNs within rapidly evolving field.

Language: Английский

---

Climate Change Versus Economic Growth: Quantifying, Identifying and Comparing Articulations in News Media Using Dynamic Topic Modeling

Environmental Communication, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 23

Published: Jan. 30, 2025

Climate change and economic growth are often portrayed as incompatible in both scientific literature the media, yet they treated separate themes. This study analyses 24 years of Finnish news data (2000–2023, n = 39,375) using dynamic topic modeling to explore how these themes intersect have evolved over time. Results reveal that climate emerged a distinct within dataset, vice versa. Initially focused on emissions international agreements, discourse coverage connect with sustainable energy practices broader socio-economic issues. Conversely, appeared discussing governmental, business, societal perspectives, critiquing capitalism emphasizing welfare education. The findings demonstrate gradual shift toward integrating environmental narratives, suggesting is increasingly viewed essential addressing change. In applying articulation theory modeling, highlights importance contextualizing machine-driven methodology results socio-political landscapes.

Language: Английский

---

Topic Mining and Evolution Trend Analysis of Fintech Research Based on the Bertopic Model

Published: Jan. 1, 2025

Language: Английский

---

Using Topic Modeling as a Semantic Technology: Examining Research Article Claims to Identify the Role of Non-Human Actants in the Pursuit of Scientific Inventions

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(6), P. 3253 - 3253

Published: March 17, 2025

Actor-network theory (ANT) represents a research paradigm that emerged within science and technology studies by explicitly focusing on the contingency of scientific inventions role non-human actants in invention course action. The article adopts an ANT perspective to focus Sub-Wavelength Grating (SWG) photonic metamaterials members group National Research Council (NRC) Canada. results are based unstructured interviews with key inventor two domain experts as well textual analysis (topic modeling) contributions novelty claims corpus articles NRC crafting concept potential applications SWGs photonics domain. Topic modeling is type statistical uses unsupervised machine learning identify clusters or groups similar words body text. It semantic structures texts understand data without predefined tags training data. Adopting topic allowed identification factors actants: (a) design simulations (b) fabrication techniques facilities used produce physical prototypes devices incorporating invented SWG waveguiding effect. Using ANT-inspired provides significant opportunities for future research.

Language: Английский

---

Machine learned text topics improve drop-out risk prediction but not symptom prediction in online psychotherapies for depression and anxiety

Psychotherapy Research, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 16

Published: March 18, 2025

Objective: Internet-delivered cognitive behavior therapies (iCBT) are effective and scalable treatments for depression anxiety. However, treatment adherence remains a major limitation that could be further understood by applying machine learning methods to during-treatment messages. We used learned topics predict drop-out risk symptom change in iCBT. Method: applied topic modeling naturalistic messages from 18,117 patients of nationwide iCBT programs generalized anxiety disorder (GAD). elastic net regression outcome predictions cross-validation aid model selection. left 10% the data as held-out test set assess predictive performance. Results: Compared reference covariates, inclusion variables resulted significant decrease prediction loss, both between-patient within-patient session-by-session models. Quantified partial pseudo-R2, increase variance explained was 2.1–6.8 percentage units. Topics did not improve compared model. Conclusions: Message contents can associated with between-patients drop-out. Our predictors were theoretically interpretable. Analysis have practical implications improved assessment allocation additional supportive interventions.

Language: Английский

---

Celebrating a decade of entrepreneurship research in JBVI (2014–2024): Taking the pulse of the past to chart the future

Journal of Business Venturing Insights, Journal Year: 2025, Volume and Issue: 23, P. e00525 - e00525

Published: April 8, 2025

Language: Английский

---

Classification of human-written and AI-generated sentences using a hybrid CNN-GRU model optimized by the spotted hyena algorithm

Alexandria Engineering Journal, Journal Year: 2025, Volume and Issue: 126, P. 116 - 130

Published: April 26, 2025

Language: Английский

---