Σχεδιασμός και υλοποίηση μοντέλου νευρωνικού δικτύου για διάγνωση άσθματος με βάση προσωπικά και υγειονομικά δεδομένα
Design and implementation of a neural network model for asthma diagnosis based on personal and health data
View/
Keywords
Νευρωνικό δίκτυο ; Μηχανική μάθηση ; Multilayer perceptron ; Άσθμα ; Μετρική απόδοσης ; ΜοντέλοAbstract
This dissertation focuses on the development and evaluation of a Multilayer Perceptron (MLP) neural network model for the prediction of asthma diagnosis. The model was implemented in Python using the scikit-learn library and was accompanied by thorough data preprocessing, including normalization, outlier detection and treatment, as well as data balancing techniques such as SMOTE and Tomek Links. Different network configurations were tested, while the Elbow Method was applied to determine the optimal number of hidden layers and neurons.
The experimental results demonstrated that the proposed model achieved a high prediction accuracy of 92%, outperforming other models reported in the literature. The study highlights the importance of proper preprocessing and parameter selection, while also confirming the potential of Artificial Intelligence and neural networks as valuable tools for supporting medical diagnosis.
Finally, future extensions of this work are suggested, such as the application of more advanced architectures (e.g., RNN, CNN) or the use of real-world clinical data for further validation and integration into healthcare systems.