This list of chapter titles covers a comprehensive learning journey through Google Cloud AI Platform, from foundational concepts to advanced techniques.
I. AI Platform Fundamentals (1-20)
- Welcome to Google Cloud AI Platform: Democratizing AI
- Introduction to Machine Learning and AI
- Understanding the AI Platform Ecosystem
- Setting up Your Google Cloud Project
- Enabling the AI Platform APIs
- Introduction to Google Cloud Console for AI
- Working with Cloud Storage for AI Data
- Introduction to Vertex AI: The Unified Platform
- Key Components of Vertex AI: Training, Prediction, and MLOps
- Understanding Vertex AI Workbench
- Creating a Vertex AI Workbench Instance
- Introduction to Notebooks for AI Development
- Working with Jupyter Notebooks in Vertex AI
- Data Exploration and Visualization with Vertex AI
- Introduction to Machine Learning Frameworks (TensorFlow, PyTorch)
- Building a Simple Machine Learning Model
- Training Your First Model on Vertex AI
- Understanding Model Training Concepts
- Evaluating Model Performance
- Deploying Your Model for Predictions
II. Model Training and Tuning (21-40)
- Working with Training Data: Formats and Best Practices
- Data Preprocessing for Machine Learning
- Feature Engineering Techniques
- Building Custom Training Jobs
- Understanding Training Configurations
- Distributed Training for Large Datasets
- Hyperparameter Tuning for Model Optimization
- Using Vertex AI Training Service
- Introduction to AutoML Training
- Automating Model Training with AutoML
- Working with Pre-trained Models
- Fine-tuning Pre-trained Models for Custom Tasks
- Transfer Learning for Efficient Model Building
- Building Models with TensorFlow
- Building Models with PyTorch
- Working with Scikit-learn on Vertex AI
- Model Explainability and Interpretability
- Understanding Feature Importance
- Visualizing Model Predictions
- Model Versioning and Management
III. Model Deployment and Prediction (41-60)
- Deploying Models for Online Prediction
- Creating Endpoints for Model Serving
- Scaling Model Deployment
- Managing Traffic Splitting for Model Updates
- A/B Testing Different Model Versions
- Monitoring Model Performance in Production
- Introduction to Batch Prediction
- Generating Predictions in Batch Mode
- Working with Prediction Requests and Responses
- Understanding Prediction Costs and Optimization
- Integrating Models with Applications
- Building a Real-time Prediction System
- Using Vertex AI Prediction Service
- Introduction to Explainable AI for Predictions
- Getting Explanations for Model Predictions
- Working with Vertex AI Endpoints
- Managing and Monitoring Endpoints
- Deploying Models to Edge Devices
- Edge AI and Model Optimization for Edge
- Building Edge Applications with Vertex AI
IV. MLOps and Workflow Automation (61-80)
- Introduction to MLOps Principles
- Automating Machine Learning Workflows
- Using Vertex AI Pipelines
- Building and Deploying Machine Learning Pipelines
- Orchestrating Machine Learning Tasks
- Managing Pipeline Runs and Artifacts
- Introduction to Vertex AI Model Registry
- Registering and Managing Models in the Registry
- Model Versioning and Release Management
- Introduction to Vertex AI Feature Store
- Creating and Managing Features in the Feature Store
- Serving Features for Model Training and Prediction
- Monitoring Feature Store Performance
- Building CI/CD Pipelines for Machine Learning
- Automating Model Training and Deployment
- Integrating with Version Control Systems (Git)
- Using Vertex AI for Experiment Tracking
- Comparing Different Model Experiments
- Reproducing Machine Learning Results
- Building a Complete MLOps Workflow
V. Advanced Topics and Integrations (81-100)
- Working with Custom Containers for Training
- Bringing Your Own Training Environment
- Advanced Hyperparameter Tuning Techniques
- Using Bayesian Optimization for Tuning
- Working with Reinforcement Learning on Vertex AI
- Introduction to Deep Learning on Vertex AI
- Building Deep Learning Models with TensorFlow and PyTorch
- Working with GPUs for Accelerated Training
- Optimizing Model Performance for GPUs
- Introduction to Kubeflow and Vertex AI
- Running Kubeflow Pipelines on Vertex AI
- Integrating Vertex AI with other Google Cloud Services
- Connecting to BigQuery for Data Access
- Using Dataflow for Data Processing
- Integrating with Cloud Functions for Serverless Workflows
- Building a Scalable Machine Learning System
- Security Best Practices for AI Platform
- Managing AI Platform Costs
- Advanced AI Platform Troubleshooting
- The Future of AI Platform and Machine Learning on Google Cloud