Imputing missing values in high-dimensional time-series data remains a significant challenge in statistics and machine learning. Although various methods have been proposed in recent years, many struggle with limitations and reduced accuracy, particularly when the missing rate is high. In this work, we present a novel f-divergence-based bidirectional generative adversarial imputation network, tf-BiGAIN, designed to address these challenges in time-series data imputation. Unlike traditional imputation methods, tf-BiGAIN employs a generative model to synthesize missing values without relying on distributional assumptions. The imputation process is achieved by training two neural networks, implemented using bidirectional modified gated recurrent units, with f-divergence serving as the objective function to guide optimization. Compared to existing deep learning-based methods, tf-BiGAIN introduces two key innovations. First, the use of f-divergence provides a flexible and adaptable framework for optimizing the model across diverse imputation tasks, enhancing its versatility. Second, the use of bidirectional gated recurrent units allows the model to leverage both forward and backward temporal information. This bidirectional approach enables the model to effectively capture dependencies from both past and future observations, enhancing its imputation accuracy and robustness. We applied tf-BiGAIN to analyze two real-world time-series datasets, demonstrating its superior performance in imputing missing values and outperforming existing methods in terms of accuracy and robustness.
22 September 2025 in Artificial Intelligence in Health
Global healthcare expenditures continue to rise, posing substantial economic challenges, particularly for low- and middle-income countries (LMICs), where resource constraints intensify the impact. Accurate forecasting, efficient resource allocation, and equitable policy development are essential to address these growing pressures. This study presents a hybrid analytical framework that integrates generative artificial intelligence (AI) with traditional econometric and machine learning models to analyze and predict trends of healthcare expenditure. Utilizing data from the World Bank and World Health Organization, we applied generative adversarial networks, hierarchical clustering, support vector machines, and autoregressive integrated moving average models to uncover spending patterns, simulate policy scenarios, and tackle disparities in health investment. Generative AI played a pivotal role by augmenting sparse and incomplete datasets, particularly from underrepresented LMICs, identifying anomalies, and generating realistic synthetic data to support robust forecasting. This enabled the development of more inclusive, equity-focused health resource planning tools. The results demonstrate improved forecasting accuracy and offer deeper insights into regional and income-based differences in expenditure trends. By combining traditional machine learning with cutting-edge generative models, this study advances a scalable, data-driven approach to support global health decision-making. Ultimately, generative AI is highlighted as a transformative enabler of equitable, informed strategies in the management of global healthcare expenditures.