pith. sign in

arxiv: 1906.08259 · v1 · pith:NUHWA2WCnew · submitted 2019-06-19 · 💻 cs.LG · nucl-th· stat.ML

Solver Recommendation For Transport Problems in Slabs Using Machine Learning

Jinzhao Chen , Japan K. Patel , Richard Vasques This is my paper

Pith reviewed 2026-05-25 20:14 UTC · model grok-4.3

classification 💻 cs.LG nucl-thstat.ML
keywords algorithmsmachineproblemslearningtransportaccelerationbestclassification
0
0 comments X

The pith

Machine learning classifiers are trained to recommend among Richardson, diffusion synthetic acceleration, and nonlinear diffusion acceleration solvers for slab transport problems by varying three problem parameters.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Neutron transport calculations in flat layers can be solved with several different numerical methods, each faster or more accurate depending on the material and geometry. The authors created many example problems by changing three input settings, ran the three solvers on each, and recorded which solver performed best. They then trained five standard machine-learning models to predict the best solver from those three settings alone. Random forest and nearest-neighbor models gave the highest accuracy on the test cases.

Core claim

random forest and K-nearest neighbors are potentially the best suited candidates for this type of classification problem.

Load-bearing premise

That the three manipulated parameters generate a representative distribution of transport scenarios whose relative solver performance can be treated as ground truth for training.

read the original abstract

The use of machine learning algorithms to address classification problems is on the rise in many research areas. The current study is aimed at testing the potential of using such algorithms to auto-select the best solvers for transport problems in uniform slabs. Three solvers are used in this work: Richardson, diffusion synthetic acceleration, and nonlinear diffusion acceleration. Three parameters are manipulated to create different transport problem scenarios. Five machine learning algorithms are applied: linear discriminant analysis, K-nearest neighbors, support vector machine, random forest, and neural networks. We present and analyze the results of these algorithms for the test problems, showing that random forest and K-nearest neighbors are potentially the best suited candidates for this type of classification problem.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit free parameters, axioms, or invented entities; the central claim rests on the unstated assumption that solver performance labels are stable and that the three parameters suffice to distinguish regimes.

pith-pipeline@v0.9.0 · 5647 in / 892 out tokens · 28747 ms · 2026-05-25T20:14:29.106028+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.