Exploring A Computationally Tractable Approach For Flexible Machine Learning In Organic Chemistry: Case Study In Nickel-Catalyzed Cross Coupling
| Author(s) | Burns, Jackson | |
| Date Accessioned | 2022-07-07T15:33:52Z | |
| Date Available | 2022-07-07T15:33:52Z | |
| Publication Date | 2022-05 | |
| Abstract | Current strategies rely on the use of hundreds of thousands of hours of computational time, brute force being required to counter the fact that running organometallic reactions to retrieve data for an ML algorithm is often prohibitively expensive. The premise of this work is to reduce the reliance on expensive simulations by improving the design of the data which the Machine Learning (ML) algorithm uses to make predictions. This would decrease the required computational time required for study, enabling rapid prototyping and larger studies. Following the implementation of this approach, this study will then seek to experimentally verify the result in a wet lab environment. | en_US |
| Advisor | Donald A. Watson | |
| Program | Chemical Engineering | |
| URL | https://udspace.udel.edu/handle/19716/31080 | |
| Publisher | University of Delaware | en_US |
| Keywords | Machine learning | en_US |
| Keywords | Organometallic coupling | en_US |
| Keywords | Organic chemistry | en_US |
| Title | Exploring A Computationally Tractable Approach For Flexible Machine Learning In Organic Chemistry: Case Study In Nickel-Catalyzed Cross Coupling | en_US |
| Type | Thesis | en_US |