Dataset size is considered a major concern in the medical domain, where lack of data is a common occurrence. This study aims to investigate the impact of dataset size on the overall performance of supervised classification models. We examined the performance of six widely-used models in the medical field, including support vector machine (SVM), neural networks (NN), C4.5 decision tree (DT), random forest (RF), adaboost (AB), and naïve Bayes (NB) on eighteen small medical UCI datasets. We further implemented three dataset size reduction scenarios on two large datasets and analyze the performance of the models when trained on each resulting dataset with respect to accuracy, precision, recall, f-score, specificity, and area under the ROC curve (AUC). Our results indicated that the overall performance of classifiers depend on how much a dataset represents the original distribution rather than its size. Moreover, we found that the most robust model for limited medical data is AB and NB, followed by SVM, and then RF and NN, while the least robust model is DT. Furthermore, an interesting observation is that a robust machine learning model to limited dataset does not necessary imply that it provides the best performance compared to other models.

Despite the rapid advances in autonomous guidance and navigation techniques for unmanned aerial vehicle (UAV) systems, there are still many challenges in finding an optimal path planning algorithm that allows outlining a collision-free navigation route from the vehicle’s current position to a goal point. The challenges grow as the number of UAVs involved in the mission increases. Therefore, this work provides a comprehensive systematic review of the literature on the path planning algorithms for multi-UAV systems. In particular, the review focuses on biologically inspired (bio-inspired) algorithms due to their potential in overcoming the challenges associated with multi-UAV path planning problems. It presents a taxonomy for classifying existing algorithms and describes their evolution in the literature. The work offers a structured and accessible presentation of bio-inspired path planning algorithms for researchers in this subject, especially as no previous review exists with a similar scope. This classification is significant as it facilitates studying bio-inspired multi-UAV path planning algorithms under one framework, shows the main design features of the algorithms clearly to assist in a detailed comparison between them, understanding current research trends, and anticipating future directions. Our review showed that bio-inspired algorithms have a high potential to approach the multi-UAV path planning problem and identified challenges and future research directions that could help improve this dynamic research area.

Delivering electronic health care (eHealth) services across multi-cloud providers to implement patient-centric care demands a trustworthy brokering architecture. Specifically, such an architecture should aggregate relevant medical information to allow informed decision-making. It should also ensure that this information is complete and authentic and that no one has tampered with it. Brokers deployed in eHealth services may fall short of meeting such criteria due to two key behaviors. The first involves violating international health-data protection laws by allowing user anonymity and limiting user access rights. Second, brokers claiming to provide trustworthy transactions between interested parties usually rely on user feedback, an approach vulnerable to manipulation by malicious users. This paper addresses these data security and trust challenges by proposing HealthyBroker, a novel, trust-building brokering architecture for multiple cloud environments. This architecture is designed specifically for patient-centric cloud eHealth services. It enables care-team members to complete eHealth transactions securely and access relevant patient data on a “need-to-know” basis in compliance with data-protection laws. HealthyBroker also protects against potential malicious behavior by assessing the trust relationship and tracking it using a neutral, tamper-proof, distributed blockchain ledger. Trust is assessed based on two strategies. First, all transactions and user feedback are tracked and audited in a distributed ledger for transparency. Second, only feedback coming from trustworthy parties is taken into consideration. HealthyBroker was tested in a simulated eHealth multi-cloud environment. The test produced better results than a benchmark algorithm in terms of data accuracy, service time, and the reliability of feedback received as measured by three malicious behavior models (naïve, feedback isolated, and feedback collective). These results demonstrate that HealthyBroker can provide care teams with a trustworthy, transparent ecosystem that can facilitate information sharing and well-informed decisions for patient-centric care.