Shapley values for feature selection: The good, the bad, and the axioms

D Fryer, I Strümke, H Nguyen - Ieee Access, 2021 - ieeexplore.ieee.org
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 …

Explainable AI for machine fault diagnosis: understanding features' contribution in machine learning models for industrial condition monitoring

E Brusa, L Cibrario, C Delprete, LG Di Maggio - Applied Sciences, 2023 - mdpi.com
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 …

Variable importance without impossible data

M Mase, AB Owen, BB Seiler - Annual Review of Statistics and …, 2024 - annualreviews.org
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 …

Inferring feature importance with uncertainties with application to large genotype data

PV Johnsen, I Strümke, M Langaas… - PLOS Computational …, 2023 - journals.plos.org
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 …

Intraday market return predictability culled from the factor zoo

S Aleti, T Bollerslev, M Siggaard - Available at SSRN 4388560, 2023 - papers.ssrn.com
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 …

Robust data valuation with weighted banzhaf values

W Li, Y Yu - Advances in Neural Information Processing …, 2024 - proceedings.neurips.cc
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 …

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 …

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 …

Game of Privacy: Towards Better Federated Platform Collaboration under Privacy Restriction

C Wu, F Wu, T Qi, Y Wang, Y Yang, Y Huang… - arXiv preprint arXiv …, 2022 - arxiv.org
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 …

Interpretability in deep learning for finance: a case study for the Heston model

D Brigo, X Huang, A Pallavicini, HSO Borde - arXiv preprint arXiv …, 2021 - arxiv.org
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 …