jhub-machine-learning's Introduction

Breast cancer deep learning 🐍

Plan

Import dataset ✔️
Data preprocessing ✔️
1. Create X and y ✔️
2. Replace datetimes with strings ✔️
3. Replace ranges with mid point ✔️
4. Handle missing values ✔️
5. Encode catergorical variables ✔️
  1. LabelEncoder ✔️
  2. OneHotEncoder ✔️
    1. Avoid the dummy variable trap ✔️
6. Feature scaling ✔️
7. Split to training and test set ✔️
Build Artificial Neural Network ✔️
1. Import Keras ✔️
2. Initialise the ANN ✔️
3. Add the input layer ✔️
4. Add the first hidden layer ✔️
5. Add the second hidden layer ✔️
6. Add the output layer ✔️
7. Compile the ANN ✔️
8. Fit the ANN to the training set ✔️
Evaluate with k-Fold Cross Evaluation ✔️
1. k-fold cross validation ✔️
2. Present results ✔️
Predict unseen data ✔️
1. Predict the test results ✔️
2. Make the confusion matrix ✔️
3. Present the confusion matrix ✔️
Improve model ✔️
Improve code ✔️
1. Tasks as functions ✔️
Comments and notes ✔️
1. Strategy explained ✔️
Run on Colab ✔️

Using Python, create the best performing neural networks algorithm you can to predict recurrence rates of breast cancer based upon the variables provided in the attached breast cancer spreadsheet.
Document:
1. the settings you tested (and rationale for the strategy you took) along the way to optimal performance.
2. a screenshot of you using the trained algorithm to make a prediction on an unseen piece of data.