Showing 1–2 of 2 results for author: Kenney, A M
-
MDI+: A Flexible Random Forest-Based Feature Importance Framework
Authors:
Abhineet Agarwal,
Ana M. Kenney,
Yan Shuo Tan,
Tiffany M. Tang,
Bin Yu
Abstract:
Mean decrease in impurity (MDI) is a popular feature importance measure for random forests (RFs). We show that the MDI for a feature $X_k$ in each tree in an RF is equivalent to the unnormalized $R^2$ value in a linear regression of the response on the collection of decision stumps that split on $X_k$. We use this interpretation to propose a flexible feature importance framework called MDI+. Speci…
▽ More
Mean decrease in impurity (MDI) is a popular feature importance measure for random forests (RFs). We show that the MDI for a feature $X_k$ in each tree in an RF is equivalent to the unnormalized $R^2$ value in a linear regression of the response on the collection of decision stumps that split on $X_k$. We use this interpretation to propose a flexible feature importance framework called MDI+. Specifically, MDI+ generalizes MDI by allowing the analyst to replace the linear regression model and $R^2$ metric with regularized generalized linear models (GLMs) and metrics better suited for the given data structure. Moreover, MDI+ incorporates additional features to mitigate known biases of decision trees against additive or smooth models. We further provide guidance on how practitioners can choose an appropriate GLM and metric based upon the Predictability, Computability, Stability framework for veridical data science. Extensive data-inspired simulations show that MDI+ significantly outperforms popular feature importance measures in identifying signal features. We also apply MDI+ to two real-world case studies on drug response prediction and breast cancer subtype classification. We show that MDI+ extracts well-established predictive genes with significantly greater stability compared to existing feature importance measures. All code and models are released in a full-fledged python package on Github.
△ Less
Submitted 4 July, 2023;
originally announced July 2023.
-
Speciation and introgression between Mimulus nasutus and Mimulus guttatus
Authors:
Yaniv Brandvain,
Amanda M. Kenney,
Lex Flagel,
Graham Coop,
Andrea L Sweigart
Abstract:
Mimulus guttatus and M. nasutus are an evolutionary and ecological model sister species pair differentiated by ecology, mating system, and partial reproductive isolation. Despite extensive research on this system, the history of divergence and differentiation in this sister pair is unclear. We present and analyze a novel population genomic data set which shows that M. nasutus "budded" off of a cen…
▽ More
Mimulus guttatus and M. nasutus are an evolutionary and ecological model sister species pair differentiated by ecology, mating system, and partial reproductive isolation. Despite extensive research on this system, the history of divergence and differentiation in this sister pair is unclear. We present and analyze a novel population genomic data set which shows that M. nasutus "budded" off of a central Californian M. guttatus population within the last 200 to 500 thousand years. In this time, the M. nasutus genome has accrued numerous genomic signatures of the transition to predominant selfing. Despite clear biological differentiation, we document ongoing, bidirectional introgression. We observe a negative relationship between the recombination rate and divergence between M. nasutus and sympatric M. guttatus samples, suggesting that selection acts against M. nasutus ancestry in M. guttatus.
△ Less
Submitted 26 October, 2013;
originally announced October 2013.
