Predicting Property Prices Using Oracle ADW and AutoML
Have you ever wanted to be able to predict the sale price of a property? In this demo, we are going to predict the sale price of Brooklyn family homes using the power of Oracle’s Autonomous Data Warehouse (ADW) in the cloud and Oracle Machine Learning (OML) notebooks. In this demo, we will be using this Kaggle dataset — Brooklyn Home Sales, 2003 to 2017. Oracle Machine Learning extends Oracle Database and enables users to augment applications and dashboards with machine learning-based intelligence. OML delivers 30+ powerful in-database machine learning algorithms with automated functionality via SQL, R and Python APIs. Below is a basic diagram that shows Oracle Machine Learning interfaces to Oracle Database.
This demo can be divided into a few different steps. First, we will need access to an OCI tenancy. Once inside of our tenancy, we are going to create an Autonomous Data Warehouse and a database schema. We are then going to create an object storage bucket where our raw CSV file with all of our data will be stored. We will then create a live table inside of the database that will automatically read the CSV file from our bucket into an ADW table. After our data is inside our ADW, we can clean our data using SQL. After our data is clean and ready for machine learning, we will use AutoML, a new component of Oracle Machine Learning that provides a no-code browser-based interface that automates the machine learning modeling process and simplifies deployment to just a few clicks, to predict the sale price of a property. We will then visualize this model in an OML notebook and invoke the model using a REST endpoint in our terminals. The architecture looks something like this.
In this demo, we will not be touching APEX and Oracle Analytics Cloud, but I left this in the architecture above for added perspective. For ease, I have divided this demo into separate blogs so that it is easier to follow:
- Prereq: Create a Free Tier Oracle Cloud Account
- Part 1 — Creating an ADW and Preparing the Data
- Part 2 — Exploring the Data & Creating a ML Model with AutoML
- Part 3 — Invoking Our AutoML Model Using the CLI
Before you begin, you will need access to an Oracle Cloud Free Tier account. If you already have access to an Oracle Cloud Free Tier account, or access to your own tenancy, proceed to Part 1. Let me know if you have any questions, and I hope you find this series helpful and fun.
