Module 8 WORK IN PROGRESS

(Work in Progress! TODO: add a page for each link (grey text))

Basement for this module:

Getting Started with Advanced Machine Learning

• Setting Up Python for ML Integration
• Instructions on setting up Python and necessary libraries for ML model training (e.g., scikit-learn).
• Collecting and Augmenting Design Data
• Steps to preprocess and augment data within Dynamo to create a dataset for ML model training.

Working with Machine Learning

• Training a Machine Learning Model
• Guide to prepare data, train, and validate an ML model using Python and scikit-learn.
• Integrating ML Models into Dynamo
• Walkthrough on embedding trained ML models within Dynamo using Python nodes for predictive design.

Examples

Case Studies and Practical Applications

• Predictive Analytics for Structural Load
• Example: Structural Load Prediction
• Energy Usage Optimization
• Example: Energy Usage Optimization

Implementing ML in Dynamo for Prediction and Optimization

• Material Cost Prediction
• Example: Material Cost Prediction
• Performance-Based Design Adjustments
• Example: Performance-Based Design Adjustments

Assignment: Data-Driven Design Enhancement

Objective: To apply advanced ML techniques to a parametric design model for enhancing design predictions and optimizations within Dynamo.

Task Description:

1. Data Collection and Augmentation:
• Generate and augment a dataset using a parametric design model in Dynamo. Focus on a specific performance metric (e.g., energy efficiency, structural integrity).
2. Model Training:
• Use scikit-learn in Python to train a predictive ML model based on the collected data. Validate the model with a test set.
3. Integration in Dynamo:
• Integrate the trained ML model into Dynamo using the Python node. Ensure the model can make real-time predictions based on user inputs.
4. Documentation and Analysis:
• Provide comprehensive documentation of data collection, model training, integration process, and analysis of prediction accuracy.
5. Deliverables:
• Dynamo file (.dyn) incorporating the ML model.
• A report (2-3 pages) summarizing the process, findings, and insights.

Evaluation Criteria:

• Completeness: Fulfillment of all specified tasks.
• Technical Accuracy: Correct training and integration of the ML model with accurate predictions.
• Documentation: Clear and detailed process documentation.
• Innovation: Creative and effective use of ML for enhancing the design process.

Timeline:

• Kickoff: Immediate post-class.
• Deadline: One week from the class date.

Resources:

• Dynamo BIM: Official Website
• Python (scikit-learn): Official Documentation
• Previous Module Materials: CEE 120C/220C Shared Library

By the end of Module 8, students will have practical experience in data augmentation, ML model training, and integration within Dynamo to create predictive, data-driven design models. This knowledge will prepare them for more advanced challenges in parametric design and optimization.