High Dimensional Human Guided Machine Learning
This addresses the problem of integrating human insights into machine learning for high-dimensional datasets, but it is incremental as it shows no clear advantage.
The paper compares XGBoost trained on human-engineered features versus directly on data, finding that human features do not outperform but are comparable, with no concrete numbers provided.
Have you ever looked at a machine learning classification model and thought, I could have made that? Well, that is what we test in this project, comparing XGBoost trained on human engineered features to training directly on data. The human engineered features do not outperform XGBoost trained di- rectly on the data, but they are comparable. This project con- tributes a novel method for utilizing human created classifi- cation models on high dimensional datasets.