An official website of the United States government
Abstract Background Measles is among the most highly infectious human diseases. By virtue of increasingly effective childhood vaccination, together with targeted supplemental immunization activities (SIAs), health authorities in the People’s Republic of China have reduced measles’ reproduction number from about 18 to 2.3. Despite substantial residual susceptibility among young adults, more in some locales than others, sustained routine childhood immunization likely would eliminate measles eventually. To support global eradication efforts, as well as expedite morbidity and mortality reductions in China, we evaluated alternative SIAs via mechanistic mathematical modelling. Methods Our model Chinese population is stratified by immune status (susceptible to measles infection; infected, but not yet infectious; infectious; and recovered or immunized), age (0, 1–4, 5–9, …, 65+ years) and location (31 provinces). Contacts between sub-populations are either empirical or a mixture of preferential and proportionate with respect to age and decline exponentially with distance between locations at age-dependent rates. We estimated initial conditions and most parameters from recent cross-sectional serological surveys, disease surveillance and demographic observations. Then we calculated the reproduction numbers and gradient of the effective number with respect to age- and location-specific immunization rates. We corroborated these analytical results by simulating adolescent and young adult SIAs using a version of our model in which the age-specific contact rates vary seasonally. Results Whereas the gradient indicates that vaccinating young adults generally is the optimal strategy, simulations indicate that a catch-up campaign among susceptible adolescent schoolchildren would accelerate elimination, with timing dependent on uptake. Conclusions These results are largely due to indirect effects (i.e. fewer infections than immunized people might otherwise cause), which meta-population models with realistic mixing are uniquely capable of reproducing accurately.
more »
« less
This content will become publicly available on April 11, 2026
Multi-concept Model Immunization through Differentiable Model Merging
Model immunization is an emerging direction that aims to mitigate the potential risk of misuse associated with open-sourced models and advancing adaptation methods. The idea is to make the released models' weights difficult to fine-tune on certain harmful applications, hence the name immunized. Recent work on model immunization focuses on the single-concept setting. However, in real-world situations, models need to be immunized against multiple concepts. To address this gap, we propose an immunization algorithm that, simultaneously, learns a single difficult initialization for adaptation methods over a set of concepts. We achieve this by incorporating a differentiable merging layer that combines a set of model weights adapted over multiple concepts.In our experiments, we demonstrate the effectiveness of multi-concept immunization by generalizing prior work's experiment setup of re-learning and personalization adaptation to multiple concepts.
more »
« less
- Award ID(s):
- 2420724
- PAR ID:
- 10596776
- Publisher / Repository:
- AAAI Press
- Date Published:
- Journal Name:
- Proceedings of the AAAI Conference on Artificial Intelligence
- Volume:
- 39
- Issue:
- 10
- ISSN:
- 2159-5399
- Page Range / eLocation ID:
- 10546 to 10554
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
After one pandemic year of remote or hybrid instructional modes, universities struggled with plans for an in-person autumn (fall) semester in 2021. To help inform university reopening policies, we collected survey data on social contact patterns and developed an agent-based model to simulate the spread of severe acute respiratory syndrome coronavirus 2 in university settings. Considering a reproduction number of R 0 = 3 and 70% immunization effectiveness, we estimated that at least 80% of the university population immunized through natural infection or vaccination is needed for safe university reopening with relaxed non-pharmaceutical interventions (NPIs). By contrast, at least 60% of the university population immunized through natural infection or vaccination is needed for safe university reopening when NPIs are adopted. Nevertheless, attention needs to be paid to large-gathering events that could lead to infection size spikes. At an immunization coverage of 70%, continuing NPIs, such as wearing masks, could lead to a 78.39% reduction in the maximum cumulative infections and a 67.59% reduction in the median cumulative infections. However, even though this reduction is very beneficial, there is still a possibility of non-negligible size outbreaks because the maximum cumulative infection size is equal to 1.61% of the population, which is substantial.more » « less
-
Federated Learning (FL) under distributed concept drift is a largely unexplored area. Although concept drift is itself a well-studied phenomenon, it poses particular challenges for FL, because drifts arise staggered in time and space (across clients). Our work is the first to explicitly study data heterogeneity in both dimensions. We first demonstrate that prior solutions to drift adaptation, with their single global model, are ill-suited to staggered drifts, necessitating multiple-model solutions. We identify the problem of drift adaptation as a time-varying clustering problem, and we propose two new clustering algorithms for reacting to drifts based on local drift detection and hierarchical clustering. Empirical evaluation shows that our solutions achieve significantly higher accuracy than existing baselines, and are comparable to an idealized algorithm with oracle knowledge of the ground-truth clustering of clients to concepts at each time step.more » « less
-
Federated Learning (FL) under distributed concept drift is a largely unexplored area. Although concept drift is itself a well-studied phenomenon, it poses particular challenges for FL, because drifts arise staggered in time and space (across clients). Our work is the first to explicitly study data heterogeneity in both dimensions. We first demonstrate that prior solutions to drift adaptation, with their single global model, are ill-suited to staggered drifts, necessitating multiple-model solutions. We identify the problem of drift adaptation as a time-varying clustering problem, and we propose two new clustering algorithms for reacting to drifts based on local drift detection and hierarchical clustering. Empirical evaluation shows that our solutions achieve significantly higher accuracy than existing baselines, and are comparable to an idealized algorithm with oracle knowledge of the ground-truth clustering of clients to concepts at each time step.more » « less
-
null (Ed.)We present a novel approach to intro-to-programming domain model discovery from textbooks using an over-generation and ranking strategy. We first extract candidate key phrases from each chapter in a Computer Science textbook focusing on intro-to-programming and then rank those concepts according to a number of metrics such as the standard tf-idf weight used in information retrieval and metrics produced by other text ranking algorithms. Specifically, we conduct our work in the context of developing an intelligent tutoring system for source code comprehension for which a specification of the key programming concepts is needed - the system monitors students' performance on those concepts and scaffolds their learning process until they show mastery of the concepts. Our experiments with programming concept instruction from Java textbooks indicate that the statistical methods such as KP Miner method are quite competitive compared to other more sophisticated methods. Automated discovery of domain models will lead to more scalable Intelligent Tutoring Systems (ITSs) across topics and domains, which is a major challenge that needs to be addressed if ITSs are to be widely used by millions of learners across many domains.more » « less
