Breast cancer (BC) disease is the most common and rapidly spreading disease across the globe. This disease can be prevented if identified early, and this eventually reduces the death rate. Machine learning (ML) is the most frequently utilized technology in research. Cancer patients can benefit from early detection and diagnosis. Using machine learning approaches, this research proposes an improved way of detecting breast cancer. To deal with the problem of imbalanced data in the class and noise, the Synthetic Minority Oversampling Technique (SMOTE) has been used. There are two steps in the suggested task. In the first phase, SMOTE is utilized to decrease the influence of imbalance data issues, and subsequently, in the next phase, data is classified using the Naive Bayes classifier, decision trees classifier, Random Forest, and their ensembles. According to the experimental analysis, the XGBoost-Random Forest ensemble classifier outperforms with 98.20% accuracy in the early detection of breast cancer.

Recommender Systems have been widely employed in information systems over the past few decades, making it easier for each user to choose their own products based on their past behaviour. Data mining tasks and visualization tools regularly use clustering techniques in the scientific and commercial arenas. It has been shown that clustering-based methods are effective and scalable to big data sets. The accuracy and coverage of clustering-based recommender systems are, however, somewhat low. In this paper, we suggest an improved multi-view clustering method for the recommendation of items in social networks to overcome these problems. To create better partitions, the artificial Bees colony optimization algorithm (ABC) is first used to improve the initial medoids’ selection. After that, users are clustered iteratively using views of both rating patterns as well as social information using multiview clustering (MVC) (i.e. trust and friendships). Ultimately, a framework is suggested for evaluating the various options. This research study suggests a novel MVC clustering approach using the ABC optimization technique. The proposed ABC-MVC algorithm’s usefulness in terms of enhancing accuracy is demonstrated by experimental findings performed on a real-world dataset and it is observed that it performs better than the pre-existing techniques and baselines.

Water is the most important resource of the Earth and is significantly utilized for agriculture, urbanization, industry, and population. This increases the demand for water; meanwhile, the climatic condition decreases the supply of it. A rise in temperature of 1 degree Celsius might dry up 20% of renewable water resources, and to circumvent the water scarcity, it is necessary to reuse, create, and consume less water without wasting it. Water desalination is the process used to reuse the used or saline water by promptly extracting the salt or unwanted minerals and producing fresh consumable water. Based on the International Desalination Association, around 300 million people rely on desalination and the people of the Middle East region rely the most upon it. Around 7% of desalination plants are located in countries such as Saudi Arabia, Bahrain, Kuwait, and the United Arab Emirates. Reverse osmosis (RO) is the relevant desalination process in this type of area however, the conventional methods include more complexities, and hence to address this issue we proposed a novel approach known as Hybrid Capuchin and Rat swarm algorithm (HCRS) for effective water desalination technology using conventional sources and renewable energy in the middle east region. Moreover, a hybrid reverse osmosis plant model is developed for identifying renewable sources such as wind and solar energy. The proposed optimization can be used to mitigate the life cycle cost and enhances the reliability of the hybrid schemes. The experiment is conducted in a MATLAB simulator and compared the results with state-of-art works over the metrics such as relative error, system cost, and reliability. Our proposed method outperforms all the other approaches.

The Internet of Things (IoT) and associated capabilities are becoming indispensable in the planning, operation, and administration of intricate systems of all sizes. High-end learning solutions that go beyond the boundaries of the problem are necessary for addressing the variety of communication concerns (compatibility, secure communication, etc.) in IoT settings. Building machine learning (ML) networks from disparate data sources is a cutting-edge practice known as Federated Learning (FL). In this article, we implement FL between edge-based servers and devices in a sparsely populated cloud to facilitate cohesive learning and the storage of critical information in smart IoT systems. FL enables collaborative training from a common model by aggregating smaller unit models via regulated edge network participants. Further, all the susceptible device’s information and sensitive message transactions are addressed via blockchain technology. Thus, a blockchain-based security mechanism is integrated to secure user privacy and facilitate widespread practical adoption. Finally, a comparison is made between the proposed model and the three best free, open-source Federated Learning models already in use (FedPD, FedProx, and FedAvg). In terms of statistical, and data heterogeneity (>70% SDI, >97% accuracy), the experimental findings suggest that the proposed model performs better than the existing techniques.