Waste Generation Prediction in Smart Cities

Abstract

One of the hardest aspects of waste generation collection occurs during working hours. Sensors in smart cities are used to track the actual conditions of waste generation. These sensors assist in the monitoring of bins usage. However, controlling a single bin is impractical and requires a significant number of resources. Another approach will be to use historical data to assist decision makers. Thus, based on historical data, this thesis proposes an appropriate and effective method by using recurrent neural network (RNN), for predicting waste generation levels in smart cities.