Crafting the Ultimate Outdoor Forecast using MySQL HeatWave

7 Views

Thanks! Share it with your friends!

You disliked this video. Thanks for the feedback!

Published Feb 26, 2025

In this presentation, I will share the development journey of my forecasting application, highlighting the use of Oracle technologies and AI to solve a common yet complex problem. You will learn about the technical underpinnings, practical applications, and the broader implications of employing AI for environmental data analysis. You can also download the application code so you can play with the technology. Then we can get outside... and stay dry!

I like to hike with minimal gear, but I don't like to get wet. Living in the Pacific Northwest makes this challenging. Using traditional forecasts, I usually either return home wet or end up carrying worthless gear. What I need is a high precision near-term hyper-local rain forecast. This presentation demonstrates how I built this solution.

I began by leveraging Oracle Cloud Infrastructure (OCI) and Oracle MySQL, utilizing the built-in machine learning capabilities of MySQL Heatwave. This made it easier to integrate weather data from nearby airports and my personal weather station through authenticated APIs. Then I applied a sophisticated Multi-layer Perceptron (MLP) neural network model.

This AI-powered forecast, accessible directly from your smartphone, offers unprecedented precision in rain predictions up to six hours ahead.