Precise and interpretable classification of autism-related behaviors is important for initial diagnosis, personalized intervention, and support arrangements. This study proposes an interpretable machine learning (ML) model using Light Gradient Boosting Machine (LightGBM) and Categorical Boosting (CatBoost) to classify behavioral patterns into four categories (normal, mild, moderate, and severe) associated with Autism Spectrum Disorder (ASD) based on a custom 377-instance survey dataset from Iraqi parents and teachers of children aged 6-12. The model observes 16 key features across communication and social interaction, repetitive behaviors, language, and adaptive skills, preprocessed via interquartile range (IQR) outlier removal, mean imputation, and K-nearest neighbors (KNN) balancing. Shapley Additive Explanations (SHAP) provide instance-level explanations, and Permutation Feature Importance (PFI) quantifies global feature importance. CatBoost had better accuracy (99. 53%) precision recall, and F1-scores (even reaching 1. 00 for some classes), completely outshining LightGBM (97. 65%). This combination of two XAI tools improves clinicians' trust and the practicality of insights, taking ASD assessment that is both accessible and transparent a step further beyond the use of black-box sensor-based models.
Background/Objectives: Early and accurate discrimination of neurological conditions, dementia, stroke and healthy aging, remains a critical clinical challenge. Electroencephalography (EEG) is a non-invasive measure of brain dynamics and entropy-based features obtained from multichannel EEG have shown strong discriminative ability. However, existing deep learning approaches do not sufficiently address the combined challenges of small clinical cohorts and high-dimensional entropy feature spaces. In this study, a novel architecture is proposed for multi-class neurological EEG classification under extreme small-sample conditions. Methods: A novel dual-branch Channel-wise Transformer and Attention-Branch Network (EEG-ChTABNet) are pr to
... Show MoreElectrochemical corrosion of hydroxyapatite (HAP) coated performance depends on various parameters like applied potential, time, thickness and sintering temperature. Thus, the optimum parameters required for the development of stable HAP coatings was found by using electrophoretic deposition (EPD) technique. This study discusses the results obtained from open circuit potential-time measurements (OCP-time), potentiodynamic polarisation and immersion tests for all alloy samples done under varying experimental conditions, so that the optimum coating parameters can be established. The ageing studies of the coated samples were carried out by immersing them in Ringer’s solution for a period of 30 days indicates the importance of stable HAP c
... Show MoreThe current research aims to diagnose the type and level of influence of social media, whose dimensions are represented by (attitudes towards social media marketing, knowledge and use of social media, influence on the Internet and social media, following up on social media) on customer behavior in its dimensions (attracting attention, raising Interest, desire creation, customer response), and the research sought to answer questions related to the research problem by testing the main hypotheses to explore correlations and influence between the two variables, and to achieve research hypothesis tests, the field study was used, as the research sample reached (135) individuals from the director and the assistant director And marketing personnel
... Show MoreThe purpose of the current article lies in determining the impact of the suggestive exercises on the development of the mental capabilities of children. The researchers used the experimental method with a single-group design, which was appropriate for the core of the current article . The study samble has been specified as children aged 4-6 years in Umm Al-Rabi'een Kindergarten, with a total of 95 children. The study samble (15 children) was randomly selected. . After the exercises were completed, the post-tests have been carried out on the sample with similar circumstances as that of pre-tests. Researchers used statistical methods in the SPSS program. After the results were presented, analyzed, and discussed, The resear
... Show MoreThis study has been performed for knowing the nutritional and chemical content of one kind chamomile tea for infant and children available in the pharmacy. The results have been showed that the percentage of essential compounds which represented with moisture, protein, fat, carbohydrate, ash and calories as 7.09%,0.01%,0.01%,92,81%, 0.08% and 371,37 Kal./100g, respectively of dry weight. Also the results have been showed that the percentage of chamomile plant extract that added to the tea as 5.74%. And the result of chemical test for effective materials in alcoholic extract showed consist Tannis, Glycosides, Flavonoids, Alkialoids,and Resins.
Metaheuristics under the swarm intelligence (SI) class have proven to be efficient and have become popular methods for solving different optimization problems. Based on the usage of memory, metaheuristics can be classified into algorithms with memory and without memory (memory-less). The absence of memory in some metaheuristics will lead to the loss of the information gained in previous iterations. The metaheuristics tend to divert from promising areas of solutions search spaces which will lead to non-optimal solutions. This paper aims to review memory usage and its effect on the performance of the main SI-based metaheuristics. Investigation has been performed on SI metaheuristics, memory usage and memory-less metaheuristics, memory char
... Show More