Abstract
Background This scoping review systematically maps externally validated machine learning (ML)-based models in cancer patient care, quantifying their performance, and clinical utility, and examining relationships between models, cancer types, and clinical decisions. By synthesizing evidence, this study identifies, strengths, limitations, and areas requiring further research. Methods The review followed the Joanna Briggs Institute's methodology, Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines, and the Population, Concept, and Context mnemonic. Searches were conducted across Embase, IEEE Xplore, PubMed, Scopus, and Web of Science (January 2014-September 2022), targeting English-language quantitative studies in Q1 journals (SciMago Journal and Country Ranking > 1) that used ML to evaluate clinical outcomes for human cancer patients with commonly available data. Eligible models required external validation, clinical utility assessment, and performance metric reporting. Studies involving genetics, synthetic patients, plants, or animals were excluded. Results were presented in tabular, graphical, and descriptive form. Results From 4023 deduplicated abstracts and 636 full-text reviews, 56 studies (2018-2022) met the inclusion criteria, covering diverse cancer types and applications. Convolutional neural networks were most prevalent, demonstrating high performance, followed by gradient- and decision tree-based algorithms. Other algorithms, though underrepresented, showed promise. Lung and digestive system cancers were most frequently studied, focusing on diagnosis and outcome predictions. Most studies were retrospective and multi-institutional, primarily using image-based data, followed by text-based and hybrid approaches. Clinical utility assessments involved 499 clinicians and 12 tools, indicating improved clinician performance with AI assistance and superior performance to standard clinical systems. Discussion Interest in ML-based clinical decision-making has grown in recent years alongside increased multi-institutional collaboration. However, small sample sizes likely impacted data quality and generalizability. Persistent challenges include limited international validation across ethnicities, inconsistent data sharing, disparities in validation metrics, and insufficient calibration reporting, hindering model comparison reliability.

Abstract
The autonomous navigation of unmanned ground vehicles (UGVs) in unstructured environments, such as agricultural or forestry settings, has been the subject of extensive research by various investigators. The navigation capability of a UGV in unstructured environments requires considering numerous factors, including the quality of data reception that allows reliable interpretation of what the UGV perceives in a given environment, as well as the use these data to control the UGV's navigation. This article aims to study different PID control algorithms to enable autonomous navigation on a robotic platform. The robotic platform consists of a forestry tractor, used for forest cleaning tasks, which was converted into a UGV through the integration of sensors. Using sensor data, the UGV's position and orientation are obtained and utilized for navigation by inputting these data into a PID control algorithm. The correct choice of PID control algorithm involved the study, analysis, and implementation of different controllers, leading to the conclusion that the Vector Field control algorithm demonstrated better performance compared to the others studied and implemented in this paper.

Abstract

Abstract
Hydroponics is an advanced agricultural technique that involves growing plants without soil. Instead, plants are cultivated in a nutrient-rich water solution that provides all the essential minerals they need to thrive, allowing plants to grow either with their roots directly in the solution or supported by inert substrates like pine bark, coconut husk fiber, and rice husk. The solid waste generated from hydroponic cultivation is valuable due to its low cost, abundance, biodegradability, and renewability. These residues are rich in lignocellulosic materials, which can be extracted and refined to produce cellulose and nanocellulose (NC). In this work, cellulose and nanocellulose were extracted from residues of coconut husk fiber and a mixture of pine bark and coconut husk fiber, used in tomato and strawberry hydroponics, respectively. The residues were ground, washed, and chemically treated to obtain cellulose and NC. The chemical process involved several stages: (i) acid treatment, alkaline treatment, and bleaching to isolate cellulose, and (ii) acid hydrolysis followed by ultrasonication to obtain NC. Both materials underwent characterization using various techniques such as TGA, DSC, XRD and FTIR-ATR, which confirmed very low levels of lignin and hemicellulose. Morphological characterization through SEM revealed the presence of micro- and nano-crystals in the cellulose and NC samples, respectively, highlighting the effectiveness of the extraction method. The high purity and quality of the extracted materials make them competitive with commercially available products, suitable for applications in healthcare, food packaging, and automotive industries, while supporting recycling and reuse principles.

Abstract
Web applications are widespread and can be accessed from anywhere, in theory, using aweb browser on a computer or smartphone. Primarily due to the diversity of web browsers and frameworks available for developing web application interfaces, testing such applications is a challenging task. With the advent of large language models, several works are utilizing them to automate software engineering tasks, including test case generation. This use of LLMs for test case generation prioritizes unit testing. More recently, we have seen the advent of Generic Artificial Intelligence Agents, which are tools that utilize LLMs and also possess the ability to run additional tools, such as cloning repositories, navigating websites, and compiling programs. In this work, which is part of a research and development project, we evaluate a specific Generic AI Agent Assistant regarding its capability to navigate web applications and create fully automated end-to-end test cases, utilizing Selenium WebDriver and JUnit 5 framework. Results show that, considering a set of nine websites, in overall end-to-end test case generation, Suna configured with DeepSeek V3 produced 165 successful test cases out of 481 generated tests, a success rate of 34.3%. On the other hand, Suna configured with Claude Sonnet 4 produced 336 successful test cases out of 479 generated tests, a success rate of 70.1%, which is very impressive, mainly due to the complexity of creating end-to-end testing. In terms of cost, we used a free and a paid LLM model. The paid model generates successful test cases at an average price of $ 0.15 per test case.

Abstract
The increasing number of connected and autonomous vehicles generates an even greater demand for efficient content delivery in vehicular networks. Estimating the popularity of content is an important task to proactively cache and distribute content throughout the networks to add value to users' experiences and reduce network congestion. This paper presents a novel approach for predicting popular content on vehicular networks based on a Federated Learning-Adversarial Autoencoder model and anonymised data. Unlike prior works that relied on users' raw features, our model protects user privacy through data anonymisation. This allows us to learn from the hidden patterns of content popularity and deliver popular content without compromising user privacy. Experiments showed that our approach exceeded traditional collaborative filtering and deep learning methods in terms of accuracy and robustness, even with sparse data.