Happy to share a new working paper: "Machine Learning Models Trained on a Longitudinal Survey Dataset Poorly Predict Hundreds of Life Outcomes" arxiv.org/abs/2610.17115 You can hear Emily Cantrell present it Thursday Oct 15 at Oxford: metrics-and-models.github.io (online parti…
Machine Learning Models Trained on a Longitudinal Survey Dataset Poorly Predict Hundreds of Life Outcomes arxiv.org
AI Weekly's analysis
→
- Across 484 life outcomes drawn from the Future of Families and Child Wellbeing Study, the median cross-validated R² was 0.03.
- The best-predicted outcomes topped out at R²CV of 0.63 for caregiver marital status and 0.62 for youth height in centimeters.
- The authors trained up to 15,160 models per outcome and still found complex procedures showed little skill over a simple baseline.
Read full analysis →
View on Bluesky ·
♥ 74
↻ 30
↩ 0
·
2 from the directory shared this ·
7d ago
Happy to share a new working paper: "Machine Learning Models Trained on a Longitudinal Survey Dataset Poorly Predict Hundreds of Life Outcomes" arxiv.org/abs/2610.17115 You can hear Emily Cantrell present it Thursday Oct 15 at Oxford: metrics-and-models.github.io (online parti…
Welcome to the Metrics and Models homepage! metrics-and-models.github.io
View on Bluesky ·
♥ 74
↻ 30
↩ 0
·
2 from the directory shared this ·
7d ago
School of Social Science at the Institute for Advanced Study call for applications for residential fellows in 2027-28. Annual theme will be: Understanding AI. Applications due Oct 15. www.ias.edu/sites/defaul...
ias.edu