Shapley values for feature selection: The good, the bad, and the axioms
The Shapley value has become popular in the Explainable AI (XAI) literature, thanks, to a
large extent, to a solid theoretical foundation, including four “favourable and fair” axioms for …
large extent, to a solid theoretical foundation, including four “favourable and fair” axioms for …
Explainable AI for machine fault diagnosis: understanding features' contribution in machine learning models for industrial condition monitoring
Although the effectiveness of machine learning (ML) for machine diagnosis has been widely
established, the interpretation of the diagnosis outcomes is still an open issue. Machine …
established, the interpretation of the diagnosis outcomes is still an open issue. Machine …
Variable importance without impossible data
The most popular methods for measuring importance of the variables in a black-box
prediction algorithm make use of synthetic inputs that combine predictor variables from …
prediction algorithm make use of synthetic inputs that combine predictor variables from …
Inferring feature importance with uncertainties with application to large genotype data
Estimating feature importance, which is the contribution of a prediction or several predictions
due to a feature, is an essential aspect of explaining data-based models. Besides explaining …
due to a feature, is an essential aspect of explaining data-based models. Besides explaining …
Intraday market return predictability culled from the factor zoo
We provide strong empirical evidence for time-series predictability of the intraday return on
the aggregate market portfolio by exploiting lagged high-frequency cross-sectional returns …
the aggregate market portfolio by exploiting lagged high-frequency cross-sectional returns …
Robust data valuation with weighted banzhaf values
Data valuation, a principled way to rank the importance of each training datum, has become
increasingly important. However, existing value-based approaches (eg, Shapley) are known …
increasingly important. However, existing value-based approaches (eg, Shapley) are known …
Efficient Shapley performance attribution for least-squares regression
L Bell, N Devanathan, S Boyd - Statistics and Computing, 2024 - Springer
We consider the performance of a least-squares regression model, as judged by out-of-
sample R 2. Shapley values give a fair attribution of the performance of a model to its input …
sample R 2. Shapley values give a fair attribution of the performance of a model to its input …
Sequential decompositions at their limit
G Junike, H Stier, MC Christiansen - arXiv preprint arXiv:2212.06733, 2022 - arxiv.org
Sequential updating (SU) decompositions are a well-known technique for creating profit and
loss (P&L) attributions, eg, for a bond portfolio, by dividing the time horizon into subintervals …
loss (P&L) attributions, eg, for a bond portfolio, by dividing the time horizon into subintervals …
Game of Privacy: Towards Better Federated Platform Collaboration under Privacy Restriction
Vertical federated learning (VFL) aims to train models from cross-silo data with different
feature spaces stored on different platforms. Existing VFL methods usually assume all data …
feature spaces stored on different platforms. Existing VFL methods usually assume all data …
Interpretability in deep learning for finance: a case study for the Heston model
Deep learning is a powerful tool whose applications in quantitative finance are growing
every day. Yet, artificial neural networks behave as black boxes and this hinders validation …
every day. Yet, artificial neural networks behave as black boxes and this hinders validation …