Artificial Intelligence (AI) and Machine Learning (ML) have become increasingly necessary for enterprises to translate today's data into direct business value. This course introduces learners to the basic concepts of AI/ML, using a mix of engaging lectures and hands-on activities to help you navigate the first steps in this exciting field. The exercises present real world use cases that show how Oracle is actively helping companies maximize business benefits through AI and ML.

This hands-on workshop focuses on Oracle Machine Learning for Python (OML4PY) which enables AI and ML capabilities of the Oracle Database in conjunction with the popular Python programming language and Jupyter notebooks. You will learn to apply machine learning concepts and algorithms to solve real-world problems ranging from data acquisition and feature engineering, to creating ML models and tuning for optimal results.

• The lab environment will be provided by Oracle on the Oracle Cloud. The students only need to bring a laptop with Microsoft Remote Desktop software preinstalled from the Microsoft Store (Windows) or Apple App Store (macOS).
• Knowledge of Python is a plus.
• No previous knowledge of machine learning required.

Lab 0: Setup the Lab Environment

This lab introduces the student lab environment and contains the steps to setup the student desktop for connecting to the lab virtual machine.

Lab 1: Introduction to Python

Get introduced to Python and Jupyter. Learn to work with basic python constructs and Pandas DataFrames objects.

Lab 2: Introduction to Oracle Machine Learning for Python

Oracle Machine Learning for Python (OML4Py) is a component of the Oracle Database Enterprise Edition that makes open source Python ready for the enterprise and big data. OML4Py integrates Python with the Oracle Database and provides a comprehensive, database-centric environment for end-to-end analytical processing. In this lab, the student will learn to use OML4Py to connect to the Oracle Database, move data between Python and the database, and store Python objects in the database.

Lab 3: Data Preparation and Exploration

Learn to perform exploratory data analysis of database-resident data and apply best practice techniques to prepare the data for machine learning. This hands-on lab highlights the OML4py Transparency Layer, and demonstrates that many familiar Python functions automatically translate to SQL and run inside the database for optimal performance and execution.

Lab 4: Embedded Python Execution

Embedded Python Execution, a feature of Oracle Machine Learning for Python, gives you the ability to invoke user-defined Python functions in Python engines that run on the Oracle Database server. Learn to store Python code in the database for reuse and sharing, and to invoke third-party Python packages for machine learning on the database server.

Lab 5: Use Regression to Estimate House Prices

Explore the CA housing prices dataset and build machine learning models for predicting house prices in the state of California. The lab walks through a real-world use case and the steps to apply an AI/ML solution for building accurate prediction models.

• Lab : Build prediction models to classify fraudulent credit card transactions

Build machine learning models to detect fraudulent credit card transactions using real-world data from financial institutions (Kaggle dataset). Prepare an unbalanced dataset using popular techniques such as under-sampling and utilize k-fold to learn iteratively and find the best hyperparameters for the algorithm. Create confusion matrix and compare prediction accuracy using Support Vector Machines, Logistic Regression and Neural Nets. You will also learn to save python objects and scripts in the Oracle database, a feature of OML4Py.

• Lab : Clustering wine data using unsupervised learning algorithms

Use unsupervised learning methods such as K-Means to cluster the Wine Data sourced from three different cultivars from Italy. Using data scaling and dimensionality reduction techniques like PCA and ICA, you will automatically generate independent set of features that help explain the variance in the dataset. Analyze the results of multiple approaches for clustering the Wine data and report accuracy.

Lab 6: Use AutoML for Automatic Feature and Model Selection

Automate end-to-end process of applying machine learning to real-world problems using the emerging AutoML techniques. In this lab you will use the automatic feature selection, model selection, and hyperparameter tuning of AutoML, a feature of OML4Py.