Application of Data Mining to Evaluate the Survival on the Titanic

• Main Authors: Vannaprapa, Noppadon, Penmetsa, Srujana, Martinez, Jesus
• Format: Text
• Language: unknown
• Published: PDXScholar 2014
• Subjects: Data mining > Algorithms; Data mining > Evaluation; Titanic (Steamship) > Disasters > Statistical aspects; Categorical Data Analysis; Survival Analysis; North Atlantic
• Online Access: https://pdxscholar.library.pdx.edu/etm_studentprojects/285; https://pdxscholar.library.pdx.edu/cgi/viewcontent.cgi?article=1284&context=etm_studentprojects

In April 1912, the largest passenger steamship in the world carrying 2229 people, the Titanic, sank after strucking an iceberg in the icy waters of the North Atlantic. In this tragic accident 1,517 people died, being one of the deadliest maritime disaster in history. The large number of deaths was due to many factors: the ship only carried enough lifeboats for 1,178 people but only 713 people survived, some of the boats didn’t deployed or had problems, and many of the lifeboats that left were not full. While children and women were prioritized to escape first, many passenger and crew member were unable to get onto any lifeboats. There were also rumors of wealthy passengers who bribed the crew to let them escape on lifeboats with a handful of survivors. With this paper, we are trying to evaluate two different data mining approaches to determine whether an individual would, or would not, survived. Our data mining algorithms use the same data set, which is based on personal information of those passengers aboard the Titanic on that fatidic day.