Bibliographic record
Abstract
Artificial Intelligence (AI) has fueled advances in many fields including healthcare, busi- ness and engineering. However, AI methods such as deep learning treat decision making as a black box where inputs are converted into outputs using a large number of parameter- ized features, with the reasoning behind decisions being largely opaque. It is difficult for humans to trust decisions when they don’t know the reasoning behind them. The goal of human-centered AI is to build reliable and safe AI systems. Human-in-the-Loop (HITL) systems build on earlier human factors approaches to complex aviation and nuclear plant interfaces, where trust is increased by integrating human supervision and expertise into the automation/AI system. Human factors engineering seeks to reduce human error, increase productivity, and enhance safety and comfort with a specific focus on the interaction be- tween the human and the automation/AI. In the research reported in this dissertation I merged human-computer interaction approaches to user experience design with human fac- tors approaches. In the research reported in this dissertation I show how human factors can be applied in the field of data-driven healthcare. I present empirical findings concerning the value of human experts in improving machine learning clinical prediction based on medi- cal data. I also report on the design and development of tools and approaches for making machine learning prediction models explainable and usable in the context of data-driven healthcare. Since highly skilled data scientists and machine learning experts will always be scarce relative to the ever increasing needs to use large data sets to improve decision mak- ing, the present research points the way towards human factors tools/systems that can allow users (such as physicians) without strong machine learning or data mining backgrounds, to use AI-based clinical decision support systems that they can trust.
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 | 0.017 | 0.077 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.004 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.005 | 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".