DeepDeath: Learning to Predict the Underlying Cause of Death with Big Data

TL;DR

DeepDeath introduces a deep learning approach using LSTMs to predict causes of death from large-scale mortality data, outperforming traditional N-gram models and capturing temporal patterns without manual feature engineering.

Contribution

The paper presents a novel deep neural network model for cause-of-death prediction that leverages LSTMs to learn temporal dependencies directly from big mortality data.

Findings

DeepDeath outperforms N-gram models in accuracy.

LSTM-based model captures temporal patterns effectively.

Both models significantly outperform random classifiers.

Abstract

Multiple cause-of-death data provides a valuable source of information that can be used to enhance health standards by predicting health related trajectories in societies with large populations. These data are often available in large quantities across U.S. states and require Big Data techniques to uncover complex hidden patterns. We design two different classes of models suitable for large-scale analysis of mortality data, a Hadoop-based ensemble of random forests trained over N-grams, and the DeepDeath, a deep classifier based on the recurrent neural network (RNN). We apply both classes to the mortality data provided by the National Center for Health Statistics and show that while both perform significantly better than the random classifier, the deep model that utilizes long short-term memory networks (LSTMs), surpasses the N-gram based models and is capable of learning the temporal aspect of the data without a need for building ad-hoc, expert-driven features.