Facing the challenge of unexpected rain during hikes in the Pacific Northwest, I developed a high-precision, near-term, hyper-local rain forecast solution. Leveraging Oracle Cloud Infrastructure, MySQL HeatWave AutoML capabilities, and real-time weather station data from Weather Underground (WU), this AI-powered forecast offers precise rain predictions up to six hours ahead, accessible via smartphone using any weather station in the WU network. This presentation outlines the development of this application, emphasizing the use of Oracle technologies and AI/ML to enhance outdoor experiences by providing reliable weather predictions, allowing hikers to stay dry.
Learning Objective 1: Understand how to leverage Oracle AI technology and airport weather station data to develop a machine learning model for forecasting rainfall.
Learning Objective 2: Gain knowledge in gathering and processing time-series weather data, and in building and evaluating a Multi-layer Perceptron (MLP) model.
Learning Objective 3:Learn how to use a REST API to access on-demand rain forecasts and how Oracle's Cloud, Oracle MySQL HeatWave, and live METAR airport weather data contribute to providing localized rain predictions.
