© 2019 IEEE. Parking is increasingly an issue in the world today especially in large and growing cities with contemporary urban mobility. The effort spent in searching for available parking spots results in significant loss of resources such as time, and fuel, as well as environmental pollution. Parking Availability can be influenced by many factors such as time of day, day of week, location, nearby events, weather and traffic conditions. Driven by the idea of predicting parking availability to help drivers plan ahead of time, we contribute a Parking Availability Forecasting Model, which uses a time-series analysis and machine-learning algorithms to predict the number of available parking spots at a certain location on a desired date and time. The forecasting model is trained on historical parking data from the cities of Kansas City, US and Melbourne, Australia. This paper also compares the accuracy of different time-series forecasting models, and how each of them fits our use-case scenario. Multivariate data analysis together with temperature and weather summary are used to cross-validate our forecasting model.