Hi, I'm Ahmad Al-shomaree

This project leverages deep learning and computer vision techniques to classify brain tumor types from medical images. The workflow begins by preprocessing a dataset containing MRI scans of four classes: glioma tumor, meningioma tumor, pituitary tumor, and normal brains. Images are loaded, resized, normalized, and labels are one-hot encoded to prepare them for model training. A convolutional neural network (CNN) architecture is built using TensorFlow/Keras, featuring multiple convolutional and pooling layers, followed by dense layers and dropout for regularization. Data augmentation is applied to the training set to improve the model's ability to generalize. The dataset is split into training, validation, and test sets, and the model is trained over several epochs, monitoring accuracy and loss. After training, the model achieves over 80% accuracy on the test set, demonstrating strong performance in multi-class classification of brain tumor images. The notebook provides an interactive interface to upload new images; the model processes these and predicts the likely tumor type, displaying the result with confidence score. Overall, this project offers a practical approach to medical image analysis, automating brain tumor detection using modern AI and computer vision methods.

This statistical study examines how air quality is influenced by several key factors, including concentrations of specific gases (CO, NO₂, and SO₂), proximity to industrial areas, and population density. The analysis also investigates the correlation between temperature and humidity in relation to these factors to determine whether they contribute to variations in air quality.

This system enables large stores and supermarkets to manage and organize their inventory efficiently. It provides clear visibility into available stock, items that are running low, storage locations, and quantity levels for each product. The system features an intuitive user interface and supports both Arabic and English. Its core functionality is structured around four main categories—clients, products, shipments, and debts—presented in a straightforward, user-friendly manner that facilitates easy understanding and operation.

This study analyzes the patterns of energy consumption in two Spanish cities—Madrid and Valencia—during the summer and winter seasons over a four-year period (2015–2019). The project also develops a model to forecast the next 24 hours of electricity demand using selected weather variables, including precipitation rate, cloud cover, and temperature. In addition, it estimates the corresponding cost of the projected energy consumption.