Foundations of Predictive AI

Software

• All attendees must have a modern web browser and an Internet connection.

Basics of Predictive AI

• Defining Machine Learning, AI, Predictive AI, and Generative AI
• Overview of Neural Networks
• The Machine Learning Workflow
• Data Collection, Exploration, and Preprocessing
• Model Training, Evaluation, and Selection
• Model Deployment and Monitoring
• Applications of Predictive AI
• Challenges in Machine Learning

Preparing Real-World Data for Use

• Data Collection & Loading
• Handling Missing Data
• Handling Duplicates
• Data Type Conversion
• Standardization & Normalization
• Handling Outliers
• Feature Engineering
• Encoding Categorical Data
• Data Integration & Merging
• Data Exporting & Storage

Supervised Learning: Classification

• Introduction to Classifiers
• Labels/“Ground Truth”
• Training a Classified with Scikit-learn
• Evaluating the Performance of a Classifier
• Generalizability and Overfitting
• Separating Training and Test Data
• Limiting Tests Performed on Test Data

Supervised Learning: Regression

• Introduction to Regression Prediction
• Linear Models
• Neural Network Regressors
• Training a Regressor with Scikit-learn
• Regularization Techniques for Linear Models
• Ridge Regression
• Lasso Regression

Supervised Learning: Ensembling Techniques

• Combining Multiple Models
• Bagging and Boosting
• Voting and Averaging
• Random Forests
• Gradient-Boosting Trees

Unsupervised Learning

• Learning from Unlabeled Data
• Clustering Techniques and Algorithms
• Gaussian Mixtures
• Dimensionality Reduction
• Reduction for Prediction vs. Visualization

Model Selection & Evaluation

• Cross-Validation to Evaluate Model Performance
• Using Scikit-Learn to Run Cross-Validation
• Hyper-Parameter Tuning
• Grid vs Random Search
• Validation and Learning Curves

Foundations of Neural Networks

• Neurons in Brain vs. Transistors in Chips
• XOR Problem (Non-Linear Functions)
• Inner Layer (Primitive Neural Network)
• Non-Linear Activation Function
• Universal Function Approximator
• Loss Functions and Gradient-Boosting

Basics of Deep Learning

• Back-Propagation and Activation Functions
• Batch and Mini-Batch Training
• Dense Layers
• GPUs and Linear Algebra
• Layers for Specific Data Types
• Deep Learning Hyper-Parameters