Frontiers in public health · 2026
Forecasting Kazakhstan's regional population aged 6-17 years to 2030: cohort validation and uncertainty analysis.
СтатьяВалидационное исследованиеНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Jainakbayev Nurlan, Orakbay Lyazat, Zhubanysheva Karlygash, Kamalieva Madina, Elzhanova Zulfiya, Baisynov Nurlan, Numanova Firuza
Аннотация
BACKGROUND: Regional forecasts of the population aged 6-17 years can provide demographic denominators for education and school health planning, but they do not directly estimate health-service or workforce requirements. This study aimed to forecast the population aged 6-17 years across Kazakhstan's 20 territorial units through 2030 and to compare the historical accuracy of a focal age-shift cohort model with ARIMA and ETS benchmarks.
METHODS: Official aggregated population data for 2015-2025 were analyzed for 20 territorial units by region, sex, locality type, and single year of age. The cohort, ARIMA, and ETS models were evaluated on the same five expanding-window one-step targets for 2021-2025. Empirical uncertainty in cohort-transition ratios was assessed using transition-year block bootstrap resampling; coefficient sensitivity and exploratory migration-effect stress tests were also examined.
RESULTS: The cohort model had the lowest rolling-origin RMSE in 17 of 20 territories, compared with two for ETS and one for ARIMA. By 2030, the largest point-estimate increases were projected for Astana, Shymkent, Almaty city, and Mangystau Region, while the largest decline was projected for North Kazakhstan Region. Empirical uncertainty intervals supported growth in eight territories and decline in two; the direction remained uncertain in ten.
CONCLUSION: Under the primary validation design, the cohort model had the lowest historical forecast error by RMSE in most territories. The projections quantify a demographic denominator; translating them into staffing, service, or infrastructure requirements requires additional data on morbidity, coverage, utilization, accessibility, productivity, and existing capacity.
METHODS: Official aggregated population data for 2015-2025 were analyzed for 20 territorial units by region, sex, locality type, and single year of age. The cohort, ARIMA, and ETS models were evaluated on the same five expanding-window one-step targets for 2021-2025. Empirical uncertainty in cohort-transition ratios was assessed using transition-year block bootstrap resampling; coefficient sensitivity and exploratory migration-effect stress tests were also examined.
RESULTS: The cohort model had the lowest rolling-origin RMSE in 17 of 20 territories, compared with two for ETS and one for ARIMA. By 2030, the largest point-estimate increases were projected for Astana, Shymkent, Almaty city, and Mangystau Region, while the largest decline was projected for North Kazakhstan Region. Empirical uncertainty intervals supported growth in eight territories and decline in two; the direction remained uncertain in ten.
CONCLUSION: Under the primary validation design, the cohort model had the lowest historical forecast error by RMSE in most territories. The projections quantify a demographic denominator; translating them into staffing, service, or infrastructure requirements requires additional data on morbidity, coverage, utilization, accessibility, productivity, and existing capacity.