Currently open challenge in Kaggle: https://www.kaggle.com/c/titanic (link to the challenge)
The challenge is to determine if a person survived in the titanic or not (from the given data in Kaggle).
The language I used for this challenge is R.
My best prediction was able to predict survival rates for the test data with 78.229% success rate.
I used feature engineering skills as well as popular models like svm (support-vector machines), glm (general linear model), decision tree(s), and random forests for these predictions.