Topic: Prediction models to tackle municipal issues using Open Data and ML algorithms
Partner: Professor Russell Greiner, Department of Computing Science, University of Alberta
Date: Fall 2015
During the Fall 2015 semester at the University of Alberta, Koosha Golmohammadi, the City’s Data Scientist, co-advised two teams of U of A students in the course Introduction to Machine Learning. Together, with Professor Russell Greiner he provided recommendations for the projects including coaching on data extraction, methodology, models performance evaluation and interpretation of project outcomes.
Students worked with datasets from the City’s Open Data Catalogue as well as other sources of data to find solutions to municipal challenges. Two student course projects used City of Edmonton data:
- Analyzing Housing Prices and Predicting Sale Values
Students: Camila Ferreira, Diogo Silva, Oleg Oleynikov, Rodolfo Wottrich, Weijie Sun and Youdong Ma.
This work presents the results of the application of various regression techniques on a training set containing information about property advertisements in Edmonton, in order to provide estimates of advertisement values of residential properties.
- Forecasting Bus Delays
Students: Yixiong(Jack) Wang, Siyang Tian, Lu Yin, Juehui Fan and Xue Rui.
In this report, we have applied 2 of the most popular machine learning techniques in this field, support vector regression (SVR) and artificial neural network (ANN), and compared their performance in bus arrival prediction. Our experiment shows that, with respect to our dataset, SVR models outperformed ANN models at most bus stops.