Titanic-Machine-Learning

By MOHAMED ALGHALY

Applied Models:

1. Logistic Regression
2. Decision Tree
3. Random Forest
4. Support Vector Machine (SVM)
5. Ada Boost
6. Gradient Boosting
7. Naive Bayes
8. K-Nearest Neighbor (KNN)

---

I have made the data preprocessing dynamic to enable flexible modeling

I implemented the transform function to clean the data

You will just have to specify any parameter to overwrite the default data cleaning as follows:

• method: how to handle missing values

• inplace: whether to clean the data as a new dataframe or into the same one

• drop_features: whether or not we want to drop useless features

• combine_rel & remove: how to handle multicollinearity

---

Attached Files

• train.csv:

  • The dataset to model (Labeled)

• test.csv:

  • The dataset to test your model on (UnLabeled)

• Titanic.ipynb:

  • The Jupyter Notebook for the project

• Titanic.html:

  • The project's report

• Predictions.csv:

  • The predections the model made on the unlabeled test data

• scenarios.png:

  • The possible scenarios to clean the data