Mohamed E. BRIKI

Mohamed E. BRIKI

Machine Learning Engineer with a Strong Interest in Deep Learning for Medical Research. Actively Pursuing PhD Opportunities.

Breast Cancer Classification: Hybrid Deep Learning Approach (Master’s Thesis)

Published:

Tools: VGG16, SVM, MLP, Random Forest, INBreast dataset, NumPy, Pandas, Scikit-learn, Keras
Date: Apr 2020 – Nov 2020
Code:[GitHub Repo]
  • Developed a breast cancer classification model combining a fine-tuned VGG-16 CNN for feature extraction and traditional ML classifiers (SVM, MLP, Random Forest) for final prediction.
  • Improved upon baseline model (AUC = 0.95) by achieving AUC = 0.98 using SVM, with further performance gains using MLP and Random Forest.
  • Trained and evaluated the model on the INBreast dataset using Google Colab; leveraged whole-image mammogram model architecture by Li Shen.