AppliedDataScience

A repository to practice data science projects.

Potato-disease-classification

• Download dataset from Kaggle

• The dataset contains images of healthy and diseased potato leaves. We are classifying the images into healthy, early disease, bad diseased using CNN.

• If you'd like to run the code, download the dataset and add it to the project folder. (potato-disease/training/PlantVillage/)

  Folder: potato-disease/

Cancer-classification

• Download dataset from Kaggle

• We are classifying a patient into having cancer with a NN and feature engineering using the dataset.

• If you'd like to run the code, download the dataset and add it to the project folder. (cancer-detection/training/)

  Folder: cancer-detection/

Document-classification

• Using transfer learning on BERT to classify documents into categories. Freezing the encoder layer and training the classifier layer, due to data limitations.

graph TD;
    Aws-Textract-->Data-Preprocessing --> Train-Custom-BERT-Model --> Inference-Classifier-Model;  

Loading