- This notebook focuses on developing ML models and for the analysis of the "Malaria Cell Images Dataset" obtained from Kaggle which can be found here.
- The aim of this project is to develop a deep learning model to reliably distinguish the numerous photos of malaria-infected and uninfected cells in the dataset.
- The notebook is divided into a number of sections, including an exploratory data analysis (EDA), model selection, data pre-processing, model construction, training, and validation, hyperparameter tweaking, model evaluation, presentation of results, and critical analysis and discussion.
- The models used in this notebook are as follows:
- A CNN model with dropout layers, data augmentation, and batch normalization
- A CNN model with dropout layers, data augmentation, batch normalization, and transfer learning (VGG16)
- A CNN model with dropout layers, data augmentation, batch normalization, and transfer learning (EfficientNetB07)
- The python verion used in this notebook is 3.6.8 on a conda environment with GPU acceleration.
- This notebook was submitted as an assignment during my course at northumbria university, kindly do not plagiarize.