This course introduces computational scientific discovery, CSD. CSD builds upon three traditions:
We will start by grounding our approach in a naively "computer-friendly" philosophy of science, and then discuss its limitations.
Next, we consider computational limitations imposed on model search by Artificial Intelligence and allied fields.
We will then discuss several scientific discoveries in both their scientific and historical context, and how the essences of their problems were abstracted into algorithm-friendly protocols.
Finally we discuss an approach to a more unifying system under development by the instructor.
I am writing a book on Computational Scientific Discovery. Preliminary chapters and lecture notes are given below.
| Week | Reading | Lecture |
| 1 |
Chapter 1: Computational Scientific Discovery, What's the Big Deal? Chapter 2: Logical Empiricism, An Attempt At A Computation-Friendly Philosophy of Science |
Lecture 1 |
| 2 | Chapter 3: Philosophy of Science After Logical Empiricism | Lecture 2 |
| 3 | Chapter 4: Computational Data, Data Structures and Reasoning in Science | Lecture 3 |
| 4 | Chapter 5: Machine Learning and Model Search | Lecture 4 |
| 5 |
Chapter 6: BACON and Related Programs Appendix A: The Plate Tectonics Paradigm |
Lecture 5a (Bacon) Lecture 5b (Plate Tectonics) |
| 6 | Chapter 7: MECHEM | Lecture 6 |
| 7 | Chapter 8: Scientific Processes and IDS | Lecture 7 |
| 8 | Lagramge (sp) and Inductive Process Modeling | Lecture 8 |
| 9 | The Scienceomatic (I) | Lecture 9 |
| 10 | The Scienceomatic (II) | Lecture 10 |
| All material is copyrighted (c) by Joseph Phillips. All rights reserved. | ||