Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · physics/0312079

Likelihood inference in the presence of nuisance parameters

N. Reid·D. A. S. Fraser
View original source ↗Read source document ↗Export metadata ↓

Abstract

We describe some recent approaches to likelihood based inference in the presence of nuisance parameters. Our approach is based on plotting the likelihood function and the $p$-value function, using recently developed third order approximations. Orthogonal parameters and adjustments to profile likelihood are also discussed. Connections to classical approaches of conditional and marginal inference are outlined.

Explore related subjects

physics.data-an

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

N. Reid, D. A. S. Fraser. 2003-12-11. Likelihood inference in the presence of nuisance parameters. https://arxiv.org/abs/physics/0312079

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Accelerating fits in high-energy physics and beyond

Data analysis in particle physics often involves optimizing a loss function that quantifies how well a theoretical model describes the observed data, in order to estimate the parameters of the model. This is referred to as fitting. As datasets grow larger and more expensive analysis methods are used, the time required to perform this parameter estimation can be a limiting factor. This work explores techniques for accelerating the fitting process, and measures their performance on a set of models representative of a variety of physics fits. We compare the speed and precision of variants of well-known optimization algorithms, namely Newton's method, BFGS, and the Migrad algorithm (commonly used in particle physics fits). We observe significant speedups when using some algorithms: most notably, some variants of BFGS are about ten times faster than Migrad, while achieving the same precision. For quasi-Newtonian algorithms like BFGS and Migrad, we find that the initial value of the Hessian matrix has a large impact on the runtime. We additionally discuss methods for accelerating the evaluation of the loss function and its derivatives, using algebraic simplifications, automatic differentiation, and modern hardware features. We explore techniques for achieving good precision with 32-bit operations, which can achieve significantly higher throughput on the GPU. Lastly we explore efficient parallelization strategies for the CPU and GPU. All of the techniques and algorithms in this paper are implemented in our new FastFit library, which was used for all of our benchmarks.

physics.data-an↗
arXiv2026

Elementwise Adaptive Random Search for High-Dimensional Constrained Continuous Global Optimization

We propose a new stochastic method for high-dimensional constrained continuous global optimization without convexity assumptions. The algorithm combines global exploration and local refinement through coordinate-wise adaptive perturbations, whose scale is proportional to the current iterate. We show that global exploration reaches any feasible region with positive probability, while local search benefits from strong concentration properties, enabling efficient decline near optima. Their combination yields convergence to a global minimizer under mild regularity conditions. Applications to LASSO sparsity improvement as well as similar BPDN (Basis Pursuit DeNoising) problems, and to neural network pruning, illustrate the practical effectiveness of the approach.

physics.data-an↗
arXiv2026

Transient Supercriticality in Hawkes Processes: Detection, Identifiability, and Apparent Criticality

Self-exciting (Hawkes) models of epidemics, earthquakes, financial markets, and social cascades are usually calibrated with a branching ratio $n$ below one. This is a condition for stationarity, not for existence: a process with $n>1$ grows exponentially but remains finite on any finite horizon. Imposing stationarity is also misleading, because transient supercritical episodes then masquerade as near-criticality. Reconstructing the latent genealogy of events, we show that a single finite record can reveal an excursion into the regime $n(t)>1$, and back, and date its onset and end. Detectability is governed by the excess $n-1$ times the square root of the number of exposed parents. Even when the memory time of the triggering is unknown, the data determine the growth rate of activity during the episode nearly as precisely as when it is known, and hence whether $n$ crossed unity, whereas the value of $n$ requires independent knowledge of the memory time. Letting the data choose between endogenous amplification and external forcing prevents confusing the two. In the 2015 Korean MERS outbreak, our reconstruction shows that transmission by hospital cohorts had collapsed by 28 May, before the incidence peak and two weeks before the instantaneous reproduction number fell below one; in real time, it establishes the collapse a few days earlier than this standard indicator. Our framework separates stationarity, non-explosion, and transient supercritical amplification, and provides a practical route for detecting short-lived endogenous instabilities in self-exciting systems.

physics.data-an↗