English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Estimating personalized effects of treatments is a complex, yet pervasiveproblem. To tackle it, recent developments in the machine learning (ML)literature on heterogeneous treatment effect estimation gave rise to manysophisticated, but opaque, tools: due to their flexibility, modularity andability to learn constrained representations, neural networks in particularhave become central to this literature. Unfortunately, the assets of such blackboxes come at a cost: models typically involve countless nontrivial operations,making it difficult to understand what they have learned. Yet, understandingthese models can be crucial -- in a medical context, for example, discoveredknowledge on treatment effect heterogeneity could inform treatment prescriptionin clinical practice. In this work, we therefore use post-hoc featureimportance methods to identify features that influence the model's predictions.This allows us to evaluate treatment effect estimators along a new andimportant dimension that has been overlooked in previous work: We construct abenchmarking environment to empirically investigate the ability of personalizedtreatment effect models to identify predictive covariates -- covariates thatdetermine differential responses to treatment. Our benchmarking environmentthen enables us to provide new insight into the strengths and weaknesses ofdifferent types of treatment effects models as we modulate different challengesspecific to treatment effect estimation -- e.g. the ratio of prognostic topredictive information, the possible nonlinearity of potential outcomes and thepresence and type of confounding.

Benchmarking Heterogeneous Treatment Effect Models through the Lens of Interpretability