This classifier will predict if you'd die on the Titanic, based on a number of features. It was able to achieve around 80% accuracy on an SVC using Pclass, Sex, Age, SibSp, Parch as features with 20% of the data on the testing set.
A better accuracy could be achieved by tweaking the classifier hyperparameters on train_classifier.py or changing the features on tools/preprocess.py. You can find an overview of the dataset here.
To run this scrips you will need sklearn, pandas, numpy and pickle.
A Django implementation of this classifier can be found here.