Payment Type Classification on Urban Taxi Big Data using Deep Learning Neural Network

Abstract

Taxi service as a reliable means of public transportation is a public need. Classification of payment types is performed on New York City Yellow Taxi Trip Open Data that considered as big data and there is a number of unlabelled data greater than the number of labeled training data was situated. We used the framework namely learning from unlabelled data (lfun) and deep learning neural network as the classifier to address the classification problem. Experimentation to find out the better performance of using lfun was conducted. We achieved the f-measure average value reaching 0.725 for classification using the lfun framework.