Completed
14:03 - Training Results and Convergence
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Dyad SciML Tutorial - Model Discovery with Universal Differential Equations
Automatically move to the next video in the Classroom when playback concludes
- 1 00:00 - Introduction to Model Discovery
- 2 00:29 - Setting Up the Component Library
- 3 00:42 - Installing Required Packages
- 4 01:21 - Creating the Simple Pot Model
- 5 01:43 - Understanding Model Components
- 6 02:08 - Heat Sources and Thermal Connections
- 7 03:05 - Running Initial Analysis
- 8 03:47 - Setting Up Julia Environment
- 9 04:16 - Comparing Simulation vs Experimental Data
- 10 05:57 - Identifying Missing Physics
- 11 06:29 - Creating the Thermal Neural Network Component
- 12 07:18 - Understanding Neural Network Parameters
- 13 08:01 - Defining UDE Inputs and Outputs
- 14 09:17 - Creating the Neural Pot Model
- 15 10:29 - Incorporating the Neural Network
- 16 11:44 - Setting Up Training Analysis
- 17 13:33 - Training the UDE Model
- 18 14:03 - Training Results and Convergence
- 19 14:28 - Analyzing Convergence Plots
- 20 14:52 - Evaluating Calibrated Simulation Results
- 21 15:50 - Validation Against Experimental Data
- 22 16:06 - Symbolic Regression for Physics Discovery
- 23 17:07 - Understanding UDE Analysis Results
- 24 17:31 - Interpreting Candidate Models
- 25 18:31 - Selecting Physically Meaningful Models
- 26 19:26 - Summary and Key Takeaways
