BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:Mathematics & Statistics Colloquium: 2025 Summer Research Pres
 entations
X-WR-TIMEZONE:Eastern Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20260518T220552Z
UID:tag:localist.com\,2008:EventInstance_51002612810731
DTSTART:20251021T150000Z
DTEND:20251021T160000Z
DESCRIPTION:“Training Artificial Intelligence Agents to Play Games” by 
 Maggie Haeussler\, Lina Mo\, and Sidney Wright\,\nAdvisor: Dr. Stephenson\
 n\nThis talk explores the application of various artificial intelligence t
 echniques in developing strategy for combinatorial games. We study a\nfami
 ly of 2-player combinatorial games played on m-by-n grids. Players alterna
 te placing their pieces on the grid\, and\, when the grid is full\,\nseque
 nces of pieces are scored based on length. Our research includes training 
 AI agents with a variety of methods including tabular learning\nand geneti
 c algorithms involving artificial neural networks. We train these agents u
 sing a variety of non-learning agents and evaluate their\nperformance agai
 nst both non-learning agents and one another to assess the quality of deci
 sion-making. Additionally\, we study this family\nof games theoretically\,
  and we determine optimal strategies in many small cases.\n\n“Modeling L
 ong-Term Seed Viability with Survival Analysis: A 30-Year Study of Pioneer
  Plants in Costa Rican\nCloud Forests” by Wyatt Snyder\, Advisors: Brian
  Yurk\, Yew-Meng Koh\n\nPioneer plants play a crucial role in rainforest e
 cosystems. However\, they grow from seeds that often wait significant time
  periods for canopy\nopenings before they can germinate. This makes the st
 udy of long-term seed viability especially important. We applied methods o
 f survival\nanalysis to a seed viability experiment to model seed viabilit
 y across time and determine the effects of microbial pathogen exposure on 
 seed\nviability for different pioneer species. Several parametric\, semipa
 rametric and nonparametric survival models were investigated\, including a
 \nnovel parametric model incorporating initial viability. Models were fit 
 to data from a 30-year seed viability experiment examining 6 pioneer\nspec
 ies native to the Cloud Forest of Monteverde\, Costa Rica. The success of 
 parametric models varied significantly from species to species\nwith some 
 fitting well with a simple exponential or Weibull and others not fitting w
 ith any standard parametric model. Fit was significantly\nenhanced by the 
 addition of an initial viability parameter across many species and pathoge
 n exposure treatment groups. These findings advance\nthe capabilities of p
 arametric models to fit a greater variety of current-status data\, particu
 larly for ecological modeling.
GEO:42.787588;-86.106797
LOCATION:VanderWerf Hall\, 104
SUMMARY:Mathematics & Statistics Colloquium: 2025 Summer Research Presentat
 ions
URL;VALUE=URI:https://calendar.hope.edu/event/mathematics-statistics-colloq
 uium-2025-summer-research-presentations-8950
CATEGORIES:Academics
CATEGORIES:Natural & Applied Sciences Division
CATEGORIES:Mathematics and Statistics
END:VEVENT
END:VCALENDAR