Journal Club

Model-free forecasting outperforms the correct mechanistic model for simulated and experimental data
Charles T. Perretti, Stephan B. Munch, and George Sugihara
PNAS March 26, 2013 110 (13) 5253-5257; https://doi.org/10.1073/pnas.1216076110, original pdf, annotated.

(...) Here we pose a rather conservative challenge and ask whether a correctly specified mechanistic model, fit with commonly used statistical techniques, can provide better forecasts than simple model-free methods for ecological systems with noisy nonlinear dynamics. (...) we found that Markov chain Monte Carlo procedures for fitting mechanistic models often converged on best-fit parameterizations far different from the known parameters. As a result, the correctly specified models provided inaccurate forecasts and incorrect inferences. In contrast, a model-free method based on state-space reconstruction gave the most accurate short-term forecasts, even while using only a single time series from the multivariate system. (...)

Hackaton

Reusability taken seriously, hands-on : Workbox.

Data Science venues and resources

Keep and eye on https://datascience.nih.gov

Discussion on the Monjoy route (more are coming)

Cloud as a service ecosystem not a VM farm

Data Platform

...