|
|
|
|
|
|
by rectangletangle
2437 days ago
|
|
|
I did some interesting work in this space. With several thousand seasons of data I used an MLP network to visualize phenotypic plasticity (how plants react to a range of environmental conditions). Accurately cleaning the input data proved to be extremely important, because there's a tremendous amount of "noise" at the individual level when dealing with living organisms, so lots of high-quality data is necessary to tease out relationships. Establishing causality was also important, considering the potential for confounding variables. It also gave me a chance to brush up on my React/front end skills, but that was more ancillary. https://sproutling.ai/blog/harvest-simulations?jm2 https://sproutling.ai/blog/growth-simulations?jm2 |
|
|