Over the weekend, I joined two popular knowledge competitions on Kaggle, Titanic - Machine Learning from Disaster and Digit Recognizer.

I was very familiar with the first one, since I had implemented the CART algorithm in college and used this competition’s dataset to compare my implementation’s performance with that of scikit-learn Decision Trees. The second competition caught my eye because of its use of the famous MNIST dataset.

I’m really excited about the Jupyter Notebooks environment offered by Kaggle. Here are my notebooks for these competitions: