arXiv subjects

Alexander Nadel

Publications and source records attributed to Alexander Nadel.

6 recordsLinked to original sources

arXiv2026

Woodelf++: A Fast and Unified Partial Dependence Plot Algorithm for Decision Tree Ensembles

Partial Dependence Plots (PDPs) visualize how changes in a single feature affect the average model prediction. They are widely used in practice to interpret decision tree ensembles and other machine learning models. Joint-PDPs extend this idea to pairs of features, revealing their combined effect. Partial Dependence Interaction Values (PDIVs) measure feature interactions. The Any-Order-PDIVs task computes these interactions for every feature subset across all rows of the dataset. We introduce Woodelf++, a unified and efficient approach for computing all these useful explainability tools on decision tree ensembles, building on Woodelf, an algorithm for efficient SHAP computation. By deriving suitable metrics over pseudo-Boolean functions, Woodelf++ can compute PDPs (exact and approximate), Joint-PDPs, and Any-Order-PDIVs in a unified framework. Our method delivers substantial complexity improvements over the state of the art, including an exponential gain for Any-Order-PDIVs. Additionally, we introduce and efficiently compute Full PDPs, which leverage the model's split thresholds to faithfully capture its behavior across all possible feature values. Woodelf++ is implemented in pure Python and supports GPU acceleration. On a dataset with 400,000 rows, Woodelf++ computes PDP and Joint-PDP up to 6x faster than the state of the art and up to five orders of magnitude faster than scikit-learn. For Any-Order-PDIVs, the gap is even larger: Woodelf++ computes all interaction values in 5 minutes, while the state of the art is estimated to require over 1,000,000 years.

cs.LG↗

arXiv2026

Backtrackable Inprocessing

We introduce Backtrackable Inprocessing (BI), a framework that enables applying inprocessing under the current trail at any decision level, at any point during incremental SAT solving. Our approach lifts the long-standing restriction that inprocessing must be performed only at the global decision level, thereby substantially increasing its potential effectiveness. We focus on three highly efficient core techniques: subsumption, self-subsuming resolution, and Bounded Variable Elimination (BVE). We show how to ensure sound backtracking in the presence of inprocessing, and demonstrate that applying BI for incremental preprocessing after propagating assumptions yields significant performance improvements on Bounded Model Checking (BMC) benchmarks from the Hardware Model Checking Competition 2017. Implemented in the Island SAT solver (IntelSAT's fork), BI enables solving $\sim$1.5$\times$ as many difficult bounds as the baseline global-level incremental preprocessor.

cs.LO↗

arXiv2026

WOODELF-HD: Efficient Background SHAP for High-Depth Decision Trees

Decision-tree ensembles are a cornerstone of predictive modeling, and SHAP is a standard framework for interpreting their predictions. Among its variants, Background SHAP offers high accuracy by modeling missing features using a background dataset. Historically, this approach did not scale well, as the time complexity for explaining n instances using m background samples included an O(mn) component. Recent methods such as Woodelf and PLTreeSHAP reduce this to O(m+n), but introduce a preprocessing bottleneck that grows as 3^D with tree depth D, making them impractical for deep trees. We address this limitation with WoodelfHD, a Woodelf extension that reduces the 3^D factor to 2^D. The key idea is a Strassen-like multiplication scheme that exploits the structure of Woodelf matrices, reducing matrix-vector multiplication from O(k^2) to O(k*log(k)) via a fully vectorized, non-recursive implementation. In addition, we merge path nodes with identical features, reducing cache size and memory usage. When running on standard environments, WoodelfHD enables exact Background SHAP computation for trees with depths up to 21, where previous methods fail due to excessive memory usage. For ensembles of depths 12 and 15, it achieves speedups of 33x and 162x, respectively, over the state-of-the-art.

cs.LG↗

arXiv2025

From Decision Trees to Boolean Logic: A Fast and Unified SHAP Algorithm

SHapley Additive exPlanations (SHAP) is a key tool for interpreting decision tree ensembles by assigning contribution values to features. It is widely used in finance, advertising, medicine, and other domains. Two main approaches to SHAP calculation exist: Path-Dependent SHAP, which leverages the tree structure for efficiency, and Background SHAP, which uses a background dataset to estimate feature distributions. We introduce WOODELF, a SHAP algorithm that integrates decision trees, game theory, and Boolean logic into a unified framework. For each consumer, WOODELF constructs a pseudo-Boolean formula that captures their feature values, the structure of the decision tree ensemble, and the entire background dataset. It then leverages this representation to compute Background SHAP in linear time. WOODELF can also compute Path-Dependent SHAP, Shapley interaction values, Banzhaf values, and Banzhaf interaction values. WOODELF is designed to run efficiently on CPU and GPU hardware alike. Available via the WOODELF Python package, it is implemented using NumPy, SciPy, and CuPy without relying on custom C++ or CUDA code. This design enables fast performance and seamless integration into existing frameworks, supporting large-scale computation of SHAP and other game-theoretic values in practice. For example, on a dataset with 3,000,000 rows, 5,000,000 background samples, and 127 features, WOODELF computed all Background Shapley values in 162 seconds on CPU and 16 seconds on GPU - compared to 44 minutes required by the best method on any hardware platform, representing 16x and 165x speedups, respectively.

cs.LG↗

arXiv2011

From Total Assignment Enumeration to Modern SAT Solver

A new framework for presenting and analyzing the functionality of a modern DLL-based SAT solver is proposed. Our approach exploits the inherent relation between backtracking and resolution. We show how to derive the algorithm of a modern SAT solver from DLL step-by-step. We analyze the inference power of Boolean Constraint Propagation, Non-Chronological Backtracking and 1UIP-based Conflict-Directed Backjumping. Our work can serve as an introduction to a modern SAT solver functionality and as a basis for future work on the inference power of a modern SAT solver and on practical SAT solver design.

cs.LO↗

arXiv2006

A Scalable Algorithm for Minimal Unsatisfiable Core Extraction

We propose a new algorithm for minimal unsatisfiable core extraction, based on a deeper exploration of resolution-refutation properties. We provide experimental results on formal verification benchmarks confirming that our algorithm finds smaller cores than suboptimal algorithms; and that it runs faster than those algorithms that guarantee minimality of the core.

cs.LO↗