Leveraging our Teacher’s Experience to Improve Machine Learning: Application to pKa Prediction
Bibliographic record
Abstract
Machine learning (ML) is gaining momentum in chemistry for the prediction of various molecular properties. However, these models are often trained on relatively scarce, sometimes low-quality data, resulting in what we describe as memorization (rather than learning) and poorly generalizable models. Aiming to revisit the way ML is practiced in chemistry, our strategy involves imparting chemistry knowledge to ML algorithms. Teachers teach chemistry with different levels of complexity in high school and graduate studies. This is due to fundamental principles being a prerequisite to understanding more advanced concepts. We posit that teaching fundamental principles to machines to predict properties, analogous to the way we teach students, will provide more accurate models. Thus, we propose to start from fundamental principles (e.g., electronegativity and inductive effect, conjugation, aromaticity) taught to students to allow them to predict properties (e.g., pKa) and provide these principles to machines to guide them to predict more advanced, yet related, properties. Based on this teaching-based approach, we developed a pKa predictor that outperforms other state-of-the-art predictors. The ML models presented herein leverage the chemists’ knowledge and qualitative principles to quantify and predict chemical properties with high performance.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.012 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".