Aythem Khairi Kareem’s research while affiliated with University of Anbar and other places

Water quality is essential for maintaining the health of ecosystems and the overall quality of life. It is crucial to monitor water quality for effective water resource management, as high-quality water suitable for domestic, drinking, irrigation, and industrial use is not always readily available. Polluted water has serious repercussions, leading to harmful environmental, human health, and infrastructure conditions. According to a United Nations (UN) report, 1.5 million people die each year due to diseases caused by polluted water. A recent study introduced a one-dimensional convolutional neural network (1D-CNN) approach for predicting water quality specifically for irrigation and drinking purposes. The prediction of water quality relies on physicochemical parameters. This study used two datasets—one for drinking water and the other for irrigation water—each with distinct features. The results show that the proposed model achieved 97.19% accuracy for irrigation and 100% accuracy for drinking water, demonstrating the model's effectiveness in accurately categorizing water samples suitable for drinking and providing robust support for decision-making processes related to potable water and irrigation.

The prediction of proton energy shows a key part in various scientific and technological studies including particle physics, medical imaging, and radiation therapy. In the last years the regression techniques study a lot trying to discover the difficult relationships among proton properties energy, needing to be studying of additional advanced techniques. This study proposed Artificial Neural Networks (ANN) as a tool for predicting types of proton energy (Max, Total and Avg). This paper proposed a comprehensive methodology for developing an ANN model for proton energy predection, covering data preparation, model architecture design, evaluation, and prediction. This proposed can ability to learn difficult models and non-linear relationships from large datasets, making them well-suited for this task. The ANN proposed achieved the highest R2 on the Total of proton energy predation of 0.96 % while the best mean_squared_error on 0.03668 for testing dataset. The results displayed the efficiency and accuracy of the proposed ANN model in predicted proton energy. The findings highlight the ability of ANNs as a tool for proton energy prediction.

The Internet of Things (IoT) has been regarded as the most critical technology due to its resource-constrained sensors transmitted via low-power wireless technologies beneath low-power lossy networks (LLNs), where the LLN has high latency and lower throughput due to its traffic patterns. The IoT possesses low-cost and low-power sensor technology, which is characterised by its low energy consumption and low latency. In this study, machine learning (ML) techniques (XGBoost, NB, and LDA) are proposed for detecting low power lossy network (RPL) attacks of routing protocol utilising multiclass of normal and four different routing attacks (Flooding Attack, Blackhole Attack, Decreased Rank Attack, and DODAG version number attack). Through experimentation and evaluation, the XGBoost classifier demonstrated superior performance, achieving an accuracy of 92.45%.

Surface-enhanced Raman spectroscopy (SERS) is a powerful technique for molecular sensing and has gained significant attention due to its high sensitivity and selectivity. SERS based on deep learning technology have been used in this study of materials, biological recognition, food safety, and intelligence. Deep learning techniques have shown tremendous potential in various scientific fields, including spectroscopy-based detection methodologies. In this study, we propose an effective deep learning approach for SERS detection using an artificial neural network (ANN). Then, the results are compared using two datasets batch1 and batch2. This study used Rhumamine 6G (R6G) as an aim molecule in this study. The experimental develops show the effectiveness of proposed ANN.

Autism Spectrum Disorder, also known as ASD, is a neurodevelopmental disease that impairs speech, social interaction, and behavior. Machine learning is a field of artificial intelligence that focuses on creating algorithms that can learn patterns and make ASD classification based on input data. The results of using machine learning algorithms to categorize ASD have been inconsistent. More research is needed to improve the accuracy of the classification of ASD. To address this, deep learning such as 1D CNN has been proposed as an alternative for the classification of ASD detection. The proposed techniques are evaluated on publicly available three different ASD datasets (children, Adults, and adolescents). Results strongly suggest that 1D CNNs have shown improved accuracy in the classification of ASD compared to traditional machine learning algorithms, on all these datasets with higher accuracy of 99.45%, 98.66%, and 90% for Autistic Spectrum Disorder Screening in Data for Adults, Children, and Adolescents respectively as they are better suited for the analysis of time series data commonly used in the diagnosis of this disorder.

In the past decade, traditional networks have been utilized to transfer data between more than one node. The primary problem related to formal networks is their stable essence, which makes them incapable of meeting the requirements of nodes recently inserted into the network. Thus, formal networks are substituted by a Software Defined Network (SDN). The latter can be utilized to construct a structure for intensive data applications like big data. In this paper, a comparative investigation of Deep Neural Network (DNN) and Machine Learning (ML) techniques that uses various feature selection techniques is undertaken. The ML techniques employed in this approach are decision tree (DT), Naïve Bayes (NB), Support Vector Machine (SVM). The proposed approach is tested experimentally and evaluated using an available NSL–KDD dataset. This dataset includes 41 features and 148,517 samples. To evaluate the techniques, several estimation measurements are calculated. The results prove that DT is the most accurate and effective approach. Furthermore, the evaluation measurements indicate the efficacy of the presented approach compared to earlier studies.