2026 Seminars

(i) Estimation of gillnet selectivity and population length frequencies

Abstract: The size selectivity of gillnets can be estimated using the catch data from experimental fishing of gangs of net panels having different mesh sizes. Gillnet selectivity curves can take a wide variety of shapes, and currently their estimation requires consideration of several different parametric forms, with right-skewed or bimodal curves typically being preferred. Here it is shown that the generalized additive model (GAM) framework provides a convenient and more flexible alternative. The GAM approach also generalizes the scope of analysis by permitting the population length frequencies of encounter to be jointly estimated as a smooth function of length. Moreover, the GAM framework allows for the inclusion of covariates such as sex or a condition index, and accommodates hierarchical sampling designs and spatial or temporal effects. Relative fishing power of the different sized meshes can also be included, notwithstanding that care with non-identifiability is required.The ease of use of the GAM approach is demonstrated on previously published lake trout data.

(ii) Gauss Is Not Mocked: less powerful multiparameter tests

Abstract: Usually it doesn't make sense to describe one test as more powerful than another -- different tests are just powerful against different alternatives. In survey statistics there are two approaches to creating multiparameter tests: weighting using the precision matrix of the parameter estimates and weighting using what would be the precision matrix of the parameter estimates with complete data. The latter are the Rao-Scott tests; the former I will call intrinsically-weighted tests. Keiran Shao's MSc thesis surprisingly found the Rao-Scott tests to be less powerful in all the examples he simulated. I will talk about whether "the Rao-Scott tests are less powerful" is (a) meaningful and (b) true. The Gauss-Markov theorem is not the answer, but it is in the vicinity of the answer.

Abstract: It this talk I will demonstrate an application that I have been working on over the last two months to help STATS 20x students understand a particular concept. The aims of this project were fairly low, but the technologies and insights I have discovered along the way – especially with respect to large language models have been extremely revealing and much larger than I could have ever anticipated. These discoveries, I hope, will let us take a hard look at what we are teaching and what we think is important. I hope to show you some really cool stuff too.

I am aiming to have a combination of talking, demonstration and some serious discussion.

This presentation has two parts. In the first, I will discuss some of the challenges of working with adaptive designs. These are experimental designs in which the next input may depend on the data collected up to that point, breaking independence between observations. As a result, the large-sample theory of adaptive designs is significantly more difficult than in the i.i.d. case, and asymptotic normality must often be verified on a case-by-case basis. I will present open problems concerning the asymptotic properties of designs currently used in the energetics industry and research, developed to test the sensitivity of explosives.

The second part of the talk focuses on applications of TabPFN. Introduced in January last year, with a major update in October, TabPFN is a foundation model for tabular data that outperforms state-of-the-art methods such as XGBoost on many regression and classification tasks. Unlike traditional machine learning approaches, it requires no fine-tuning or additional training: TabPFN is pre-trained on an enormous corpus of synthetic datasets designed to capture nonlinear relationships in tabular problems. In addition to outlining ideas for future use, I will present one application in which we use TabPFN to improve estimates of conditional Shapley values in explainable AI.

About the speaker: Dennis Christensen is a researcher at the Norwegian Defence Research Establishment (FFI), visiting the University of Auckland from early January until mid-April. His research focuses on statistical aspects of sensitivity testing of energetic materials., with particular focus on explosive remnants of war and dumped ammunition. He completed his PhD at the University of Oslo in 2024.

Abstract: We propose a novel modeling framework for time-evolving networks allowing for long-term dependence in network features that update in continuous time. Dynamic network growth is functionally parameterized via the conditional intensity of a marked point process. This characterization enables flexible modeling of both the time of updates and the network updates themselves, dependent on the entire left-continuous sample path. We propose a path-dependent nonlinear marked Hawkes process as an expressive platform for modeling such data; its dynamic mark space embeds the time-evolving network. We establish stability conditions, demonstrate simulation and subsequent feasible likelihood-based inference through numerical study, and illustrate the methodology with an application to conference attendee social network data. The resulting methodology serves as a general framework that can be readily adapted to a wide range of network topologies and point process model specifications.