Research into decision trees to make sure they will fit into our needs, and run a small model if possible.
Define progress:
[x] Why pick decision trees over the random forest? Find reasons.
[x] Define decision trees work better in 27th patient's data than linear regress.
[x] Start considering and building hand-drawing trees to simulate decision tree behavior and understand the potential output from our data.
Sucess outcomes:
[x] Discover that non-linear is the opposite of experimenting with linear regression. Discover the 27th patient's data can be worked and treated as 27's different non-linear relations. A small model has proven it.
[x] Discover the decision trees implementation was okay in the proving stage. It requires more data and more work to tunning the implementation and level of complexity to predict better results.
[x] It was discovered decision trees do similar behavior to human intuition, which decided outcome can be done similarly by implementing "max" and "min" functions to find the general location and discover the relative region of the 27th patient's highest knee stress. Therefore, much more work has to be done to work smart with decision trees to go above and beyond than relative region with the highest stress. Instead, the model has to be intelligence enough to define the difference between prediction and max function.
Research into decision trees to make sure they will fit into our needs, and run a small model if possible.
Define progress:
Sucess outcomes: