Optimization by Adaptive Stochastic Descent

Kerr, Cliff C.; Durá-Bernal, Salvador; Smolinski, Tomasz G.; Chadderdon, George L.; Wilson, David

doi:10.1371/journal.pone.0192944

Cited by 24 publications

(26 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This second step is repeated (on subsequent iterations, the choice of program based on probability distributions learned from previous iterations) until the solution converges. Further details are provided in [ 32 ].…”

Section: Methodsmentioning

confidence: 99%

Getting it right when budgets are tight: Using optimal expansion pathways to prioritize responses to concentrated and mixed HIV epidemics

et al. 2017

Self Cite

View full text Add to dashboard Cite

BackgroundPrioritizing investments across health interventions is complicated by the nonlinear relationship between intervention coverage and epidemiological outcomes. It can be difficult for countries to know which interventions to prioritize for greatest epidemiological impact, particularly when budgets are uncertain.MethodsWe examined four case studies of HIV epidemics in diverse settings, each with different characteristics. These case studies were based on public data available for Belarus, Peru, Togo, and Myanmar. The Optima HIV model and software package was used to estimate the optimal distribution of resources across interventions associated with a range of budget envelopes. We constructed “investment staircases”, a useful tool for understanding investment priorities. These were used to estimate the best attainable cost-effectiveness of the response at each investment level.FindingsWe find that when budgets are very limited, the optimal HIV response consists of a smaller number of ‘core’ interventions. As budgets increase, those core interventions should first be scaled up, and then new interventions introduced. We estimate that the cost-effectiveness of HIV programming decreases as investment levels increase, but that the overall cost-effectiveness remains below GDP per capita.SignificanceIt is important for HIV programming to respond effectively to the overall level of funding availability. The analytic tools presented here can help to guide program planners understand the most cost-effective HIV responses and plan for an uncertain future.

show abstract

Section: Methodsmentioning

confidence: 99%

Getting it right when budgets are tight: Using optimal expansion pathways to prioritize responses to concentrated and mixed HIV epidemics

et al. 2017

Self Cite

View full text Add to dashboard Cite

show abstract

“…This process is repeated multiple times using a Monte Carlo initialization to increase the liklihood of locating the global minimum. To perform the optimization, we employ a Bayesian adaptive locally linear stochastic descent algorithm [ 17 ].…”

Section: Methodsmentioning

confidence: 99%

In the interests of time: improving HIV allocative efficiency modelling via optimal time‐varying allocations

Shattock

Kerr

Stuart

et al. 2016

Journal of the International AIDS Society

View full text Add to dashboard Cite

IntroductionInternational investment in the response to HIV and AIDS has plateaued and its future level is uncertain. With many countries committed to ending the epidemic, it is essential to allocate available resources efficiently over different response periods to maximize impact. The objective of this study is to propose a technique to determine the optimal allocation of funds over time across a set of HIV programmes to achieve desirable health outcomes.MethodsWe developed a technique to determine the optimal time-varying allocation of funds (1) when the future annual HIV budget is pre-defined and (2) when the total budget over a period is pre-defined, but the year-on-year budget is to be optimally determined. We use this methodology with Optima, an HIV transmission model that uses non-linear relationships between programme spending and associated programmatic outcomes to quantify the expected epidemiological impact of spending. We apply these methods to data collected from Zambia to determine the optimal distribution of resources to fund the right programmes, for the right people, at the right time.Results and discussionConsidering realistic implementation and ethical constraints, we estimate that the optimal time-varying redistribution of the 2014 Zambian HIV budget between 2015 and 2025 will lead to a 7.6% (7.3% to 7.8%) decrease in cumulative new HIV infections compared with a baseline scenario where programme allocations remain at 2014 levels. This compares to a 5.1% (4.6% to 5.6%) reduction in new infections using an optimal allocation with constant programme spending that recommends unrealistic programmatic changes. Contrasting priorities for programme funding arise when assessing outcomes for a five-year funding period over 5-, 10- and 20-year time horizons.ConclusionsCountries increasingly face the need to do more with the resources available. The methodology presented here can aid decision-makers in planning as to when to expand or contract programmes and to which coverage levels to maximize impact.

show abstract

“…However, since a single model run returns a scalar log-likelihood value, these runs can be easily integrated into standardized calibration frameworks. An example implementation using Weights & Biases ( wandb.com ) is included in the codebase, but any standard optimization library -such as the optimization module of SciPy -can be easily adapted, as can more advanced methods such as the adaptive stochastic descend method of the Sciris library (Kerr et al, 2018), or Bayesian approaches such as history matching (Andrianakis et al, 2015).…”

Section: Calibrationmentioning

confidence: 99%

Covasim: an agent-based model of COVID-19 dynamics and interventions

Kerr

Stuart

Mistry

et al. 2020

Preprint

Self Cite

144

192

View full text Add to dashboard Cite

The COVID-19 pandemic has created an urgent need for models that can project epidemic trends, explore intervention scenarios, and estimate resource needs. Here we describe the methodology of Covasim , an open-source model developed to help address these questions. Covasim includes demographic information on age structure and population size; realistic transmission networks in different social layers, including households, schools, workplaces, and communities; age-specific disease outcomes; and intrahost viral dynamics, including viral-load-based transmissibility. Covasim also supports an extensive set of interventions, including non-pharmaceutical interventions, such as physical distancing, hygiene measures, and protective equipment; and testing interventions, such as symptomatic and asymptomatic testing, isolation, contact tracing, and quarantine. These interventions can incorporate the effects of delays, loss-to-follow-up, micro-targeting, and other factors. In collaboration with local health agencies and policymakers, Covasim has already been applied to examine disease dynamics and policy options in Africa, Europe, Oceania, and North America.

show abstract

Optimization by Adaptive Stochastic Descent

Cited by 24 publications

References 47 publications

Getting it right when budgets are tight: Using optimal expansion pathways to prioritize responses to concentrated and mixed HIV epidemics

Getting it right when budgets are tight: Using optimal expansion pathways to prioritize responses to concentrated and mixed HIV epidemics

In the interests of time: improving HIV allocative efficiency modelling via optimal time‐varying allocations

Covasim: an agent-based model of COVID-19 dynamics and interventions

Contact Info

Product

Resources

About