A Systematic Review of Synthetic Data Generation Techniques Using Generative AI

Electronics, Journal Year: 2024, Volume and Issue: 13(17), P. 3509 - 3509

Published: Sept. 4, 2024

Synthetic data are increasingly being recognized for their potential to address serious real-world challenges in various domains. They provide innovative solutions combat the scarcity, privacy concerns, and algorithmic biases commonly used machine learning applications. preserve all underlying patterns behaviors of original dataset while altering actual content. The methods proposed literature generate synthetic vary from large language models (LLMs), which pre-trained on gigantic datasets, generative adversarial networks (GANs) variational autoencoders (VAEs). This study provides a systematic review techniques that can be identify limitations suggest future research areas. findings indicate these technologies specific types, they still have some drawbacks, such as computational requirements, training stability, privacy-preserving measures limit usability. Addressing issues will facilitate broader adoption generation across disciplines, thereby advancing data-driven solutions.

Language: Английский

---

Artificial intelligence for image recognition in diagnosing oral and oropharyngeal cancer and leukoplakia

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 29, 2025

Visual diagnosis is one of the key features squamous cell carcinoma oral cavity (OSCC) and oropharynx (OPSCC), both subsets head neck (HNSCC) with a heterogeneous clinical appearance. Advancements in artificial intelligence led to Image recognition being introduced recently into large language models (LLMs) such as ChatGPT 4.0. This exploratory study, for first time, evaluated application image by diagnose leukoplakia based on images, images without any lesion control group. A total 45 were analyzed, comprising 15 cases each SCC, leukoplakia, non-lesion images. 4.0 was tasked providing most likely these scenario one. In two history provided, whereas three only given. The results accuracy LLM rated independent reviewers overall performance using modified Artificial Intelligence Performance Index (AIPI. this demonstrated ability correctly identify alone, while SCC insufficient, but improved including prompt. Providing resulted misclassification some cases. Oral lesions more be diagnosed correctly. study lesions, convincing detecting when added, Leukoplakia detected solely recognition. therefore currently insufficient reliable OPSCC OSCC diagnosis, further technological advancements may pave way use setting.

Language: Английский

---

Assessing the use of the novel tool Claude 3 in comparison to ChatGPT 4.0 as an artificial intelligence tool in the diagnosis and therapy of primary head and neck cancer cases

European Archives of Oto-Rhino-Laryngology, Journal Year: 2024, Volume and Issue: 281(11), P. 6099 - 6109

Published: Aug. 7, 2024

Head and neck squamous cell carcinoma (HNSCC) is a complex malignancy that requires multidisciplinary tumor board approach for individual treatment planning. In recent years, artificial intelligence tools have emerged to assist healthcare professionals in making informed decisions. This study investigates the application of newly published LLM Claude 3 Opus compared currently most advanced ChatGPT 4.0 diagnosis therapy planning primary HNSCC. The results were conventional board; (2) Materials Methods: We conducted March 2024 on 50 consecutive head cancer cases. diagnostics MDT recommendations each patient rated by two independent reviewers following parameters: clinical recommendation, explanation, summarization addition Artificial Intelligence Performance Instrument (AIPI); (3) Results: this study, achieved better scores diagnostic workup patients than provided involving surgery, chemotherapy, radiation therapy. terms recommendations, explanation scored similar 4.0, listing which congruent with MDT, but failed cite source information; (4) Conclusion: first analysis cases demonstrates superior performance HNSCC recommendations. marks advent launched AI model may be assessment setting.

Language: Английский

---

An innovative GPT-based open-source intelligence using historical cyber incident reports

Natural Language Processing Journal, Journal Year: 2024, Volume and Issue: 7, P. 100074 - 100074

Published: April 24, 2024

Language: Английский

---

ChatGPT in education: benefits and challenges of ChatGPT for mathematics and science teaching practices

International Journal of Mathematical Education in Science and Technology, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 30

Published: May 28, 2024

This study investigates the utilisation and perceptions of artificial intelligence (AI) applications among mathematics science teachers in enhancing students' learning experience classrooms. One prominent AI application this research seeks to explore, given its recent rise popularity usage, is ChatGPT. The aims explore frequency purpose ChatGPT potential for student engagement, implementation's perceived effectiveness, benefits, challenges. A sample 109 respondents was surveyed gather insights into usage patterns findings shed light on how utilise ChatGPT, including generating examples, assessing difficulty, providing explanations, supporting problem-solving, test preparation. Additionally, reveals that shows promise engagement understanding. However, highlighted challenges such as accuracy, connectivity, language limitations. These contribute our understanding role technologies, specifically educational settings provide valuable effective integration teaching practices various course contents topics.

Language: Английский

---

Enhancing Multilingual Hate Speech Detection: From Language-Specific Insights to Cross-Linguistic Integration

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 121507 - 121537

Published: Jan. 1, 2024

Language: Английский

---

Radar-Based Target Tracking Using Deep Learning Approaches with Unscented Kalman Filter

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(18), P. 8332 - 8332

Published: Sept. 16, 2024

Machine learning, a rapidly growing field, has attracted numerous researchers for its ability to automatically learn from and make predictions based on data. This manuscript presents an innovative approach estimating the covariance matrix of noise in radar measurements target tracking, resulting collaborative efforts. Traditionally, have assumed that sonar is present vast majority literature related tracking. On other hand, this research aims estimate it by employing deep learning algorithms with noisy range, bearing, elevation sensors. approach, involving multiple disciplines, provides more precise accurate estimate. Additionally, unscented Kalman filter was combined gated recurrent unit, multilayer perceptron, convolutional neural network, long short-term memory accomplish task 3D tracking airborne environment. The quantification results achieved through use Monte Carlo simulations, which demonstrated network performed better than any approach. system simulated using Python program, proposed method offers higher accuracy faster convergence time conventional methods. demonstration potential collaboration can research.

Language: Английский

---

Evolving AI-Driven Workflow Management, Part B: Non-Unique Engineering Workflows and Scalable Open-weight Agents

AIAA SCITECH 2022 Forum, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 3, 2025

Language: Английский

---

Entropy-Optimized Dynamic Text Segmentation and RAG-Enhanced LLMs for Construction Engineering Knowledge Base

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(6), P. 3134 - 3134

Published: March 13, 2025

In the field of construction engineering, there exists a dynamic evolution extensive technical standards and specifications (e.g., GB/T ISO series) that permeate entire lifecycle design, construction, operation–maintenance. These require continuous version iteration to adapt technological innovations. Engineers specialized knowledge bases assist in understanding updating these standards. The advancement large language models (LLMs) Retrieval-Augmented Generation (RAG) technologies provides robust support for constructing domain-specific bases. This study developed tested vertical domain base scheme based on RAG architecture LLMs, comprising three critical components: entropy-optimized text segmentation (EDTS), vector correlation-based chunk ranking, iterative optimization prompt engineering. employs an EDTS method ensure information clarity predictability within limited lengths, followed by selecting 10 relevant chunks form prompts input into thereby enabling efficient retrieval knowledge. Experimental validation using Qwen-series LLMs with test set 101 expert-verified questions from Chinese industry standard demonstrates overall accuracy reaches 76%. comparative experiments across model scales (1.5B, 3B, 7B, 14B, 32B, 72B) quantitatively reveal relationship between size, answer accuracy, execution time, providing decision-making guidance computational resource-accuracy tradeoffs engineering practice.

Language: Английский

---

Exploring Vacuum Erection Device User Feedback: A GPT-4 Thematic Analysis

Urology, Journal Year: 2025, Volume and Issue: unknown

Published: March 1, 2025

Language: Английский

---