Part 1:  Using Oracle Analytics Cloud and Oracle Machine Learning to Explore our ADW Data and Build our Initial ML Models

What demographics, lifestyles, and habits do the Tesla audience share? How do your credit score, age, and personal habits help predict your likelihood to purchase one of these exciting electric vehicles?  This Tesla scenario includes processing structured, spatial, graph, and unstructured data to create a complete 360-degree customer view. We will take a deeper and fun look at our current Tesla owners using demo data and Oracle Autonomous Database, Oracle Machine Learning, and Oracle Analytics Cloud. We build and apply Oracle Machine Learning models inside Oracle Database to identify those people who are most likely to buy Teslas. 

This interesting and fun demo and workshop walks readers through the steps to build machine learning models using Oracle Machine Learning and highlights a few of OML’s unique differentiators including the ability to process structured unstructured, relationship, and location data all while processing the data 100% inside Oracle Autonomous Database. Then we will interactively explore our results using Oracle Analytics Cloud.

See how easy machine learning can be and whether a Tesla is likely to be in your future.

See “Is a Tesla in Your Future? A Machine Learning Approach using Structured, Unstructured, Relationship and Location Data” presentation.

Download the on-premises version presentation that uses Oracle Data Miner (click to download ODMr workflow) and Oracle Analytics Cloud.

  1. First, let’s look at our Tesla customer data using OAC, ADW and OML Notebooks

  2. Let’s do that again using OML AutoML UI

  3. Now, let’s add 3 different data types to create a more complete 360-degree customer POV:
    – Tesla reviews (unstructured data)
    – Tesla friends (relationship / “graph” data)
    – Telsa location data (geospatial / transactional location data)
  4. We build, evaluate, and apply OML models on this 360-degree customer data to target people likely to buy a Tesla

  5. Finally, we import OML models into Oracle Analytics Cloud, create a Data Flow, and interactively analyze our prospects


We start by reviewing our Tesla customers data using Oracle Analytics Cloud

We can visualize our data and explore some hunches, but to best target our likely Tesla electric car buyers, we want to use Oracle Machine Learning, a no-license required feature of Oracle Databases. Our data is managed in an Oracle Autonomous Database so we leverage an OML4SQL Classification OML Example Notebook to create an OML Notebook to build, evaluate, select and apply and OML in-database ML model to target likely Tesla buyers. 

After exploring our data again using Oracle Machine Learning Notebook’s data visualizations we build an OML Attribute Importance model to identify the key attributes in our customer data that are most influence whether a customer purchases a Tesla car or not.

We see that CREDIT_CARD_DEBT and the number of friends with electric cars (NO_FRIENDS_ELEC_CARS) and FICO_CREDIT_SCORE are the most influential attributes but we need more detail and insights.  We continue further and build and compare multiple Oracle Machine Learning models.   Notice that each model is built inside the Oracle Database in just seconds!

Now we compare the OML models and decide to select the OML random forest model to make our initial predictions.  

Now, let’s use our OML model to see who might be buying a new Tesla car.

Is a Tesla in Your Future?” Part 2: Let’s do that again using OML AutoML UI

Download the fun Tesla Demo Artifacts (Tesla demo datasets TESLA_78, TESLA_REVIEWS89, TESLA_FRIENDS89, TESLA_LOCATIONS89, Is a Tesla in Your Future_ A 360-degree POV using OML OML notebook, and Tesla Buyers OAC Project.dva)

– Naman Mehta Principal Member of Technical Staff who worked diligently to get all the complex SQL joins, text mining components to work and for his passion and help to create this fun and educational demo scenario.
– Philippe Lions, Senior Director, Product Management, Analytics Platform, Oracle Analytics Cloud (OAC) for his expertise, ideas, and help in highlighting OML working with Oracle Analytics Cloud.
– Siddesh Chikkanayakanahalli Prabhu Dev Ujjni, Staff Cloud Engineer for his dedication, hard work, knowledge, and help in creating this and other OML demo scenarios.


Source link

Leave a Reply

Your email address will not be published.