More Like this

1. Tree++: Truncated Tree Based Graph Kernels

   https://doi.org/10.1109/TKDE.2019.2946149  

   Ye, Wei ; Wang, Zhen ; Redberg, Rachel ; Singh, Ambuj ( April 2021 , IEEE Transactions on Knowledge and Data Engineering)
2. Correcting tree-ring δ13C time series for tree-size effects in eight temperate tree species

   https://doi.org/10.1093/treephys/tpz138  

   Vadeboncoeur, Matthew A ; Jennings, Katie A ; Ouimette, Andrew P ; Asbjornsen, Heidi ( January 2020 , Tree Physiology)

   Cernusak, Lucas (Ed.)

   Abstract Stable carbon isotope ratios (δ13C) in tree rings have been widely used to study changes in intrinsic water-use efficiency (iWUE), sometimes with limited consideration of how C-isotope discrimination is affected by tree height and canopy position. Our goals were to quantify the relationships between tree size or tree microenvironment and wood δ13C for eight functionally diverse temperate tree species in northern New England and to better understand the physical and physiological mechanisms underlying these differences. We collected short increment cores in closed-canopy stands and analyzed δ13C in the most recent 5 years of growth. We also sampled saplings in both shaded and sun-exposed environments. In closed-canopy stands, we found strong tree-size effects on δ13C, with 3.7–7.2‰ of difference explained by linear regression vs height (0.11–0.28‰ m−1), which in some cases is substantially stronger than the effect reported in previous studies. However, open-grown saplings were often isotopically more similar to large codominant trees than to shade-grown saplings, indicating that light exposure contributes more to the physiological and isotopic differences between small and large trees than does height. We found that in closed-canopy forests, δ13C correlations with diameter at breast height were nonlinear but also strong, allowing a straightforward procedure to correct tree- or stand-scale δ13C-based iWUE chronologies for changing tree size. We demonstrate how to use such data to correct and interpret multi-decadal composite isotope chronologies in both shade-regenerated and open-grown tree cohorts, and we highlight the importance of understanding site history when interpreting δ13C time series. 

   more » « less
3. The “AI + R” - tree: An Instance-optimized R - tree

   https://doi.org/10.1109/MDM55031.2022.00023  

   Abdullah-Al-Mamun, Abdullah-Al- ; Haider, Ch. Md. ; Wang, Jianguo ; Aref, Walid G. ( June 2022 , IEEE)
4. Estimating tree phenology from high frequency tree movement data

   https://doi.org/10.1016/j.agrformet.2018.08.020  

   Gougherty, Andrew V. ; Keller, Stephen R. ; Kruger, Anton ; Stylinski, Cathlyn D. ; Elmore, Andrew J. ; Fitzpatrick, Matthew C. ( December 2018 , Agricultural and Forest Meteorology)

   null (Ed.)
5. Tree Space Prototypes: Another Look at Making Tree Ensembles Interpretable

   https://doi.org/3412815.3416893  

   Tan, Sarah ; Soloviev, Matvey ; Hooker, Giles ; Wells, Martin ( October 2020 , Foundations of data science)

   null (Ed.)

   Ensembles of decision trees perform well on many problems, but are not interpretable. In contrast to existing approaches in interpretability that focus on explaining relationships between features and predictions, we propose an alternative approach to interpret tree ensemble classifiers by surfacing representative points for each class -- prototypes. We introduce a new distance for Gradient Boosted Tree models, and propose new, adaptive prototype selection methods with theoretical guarantees, with the flexibility to choose a different number of prototypes in each class. We demonstrate our methods on random forests and gradient boosted trees, showing that the prototypes can perform as well as or even better than the original tree ensemble when used as a nearest-prototype classifier. In a user study, humans were better at predicting the output of a tree ensemble classifier when using prototypes than when using Shapley values, a popular feature attribution method. Hence, prototypes present a viable alternative to feature-based explanations for tree ensembles. 

   more » « less