Predictive modelling for IoT enabled waste management system
Kugblenu, Carl Makafui (2018)
Diplomityö
Kugblenu, Carl Makafui
2018
School of Engineering Science, Tietotekniikka
Kaikki oikeudet pidätetään.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2018110647396
https://urn.fi/URN:NBN:fi-fe2018110647396
Tiivistelmä
The research presents a web based Decision Support System(DSS) that operates within the waste management and IoT ecosystem. The DSS leverages temperature, humidity, fill-level and weight about the Smart Garbage Bins(SGB) to improve waste collection and reduce the number of messages being sent by the SGBs by predicting the send intervals based on frequency of the same data. We demonstrate the efficiency of our approach with a decision tree and compare with four machine learning predictive classifiers: i) Random Forest ii) Naïve Bayes iii)K Nearest Neighbour iv) Support Vector Classifier. Results demonstrate that decision perform best based on accuracy, precision and recall. It also improves the waste collection process and can reduce the number of messages sent by the SGB by 44% to 45% of the original message count.