Change-point detection is a challenging problem that has a number of applications across various real-world domains. The primary objective of CPD is to identify specific time points where the underlying system undergoes transitions between different states, each characterized by its distinct data distribution. Precise identification of change points in time series omics data can provide insights into the dynamic and temporal characteristics inherent to complex biological systems. Many change-point detection methods have traditionally focused on the direct estimation of data distributions. However, these approaches become unrealistic in high-dimensional data analysis. Density ratio methods have emerged as promising approaches for change-point detection since estimating density ratios is easier than directly estimating individual densities. Nevertheless, the divergence measures used in these methods may suffer from numerical instability during computation. Additionally, the most popular -relative Pearson divergence cannot measure the dissimilarity between two distributions of data but a mixture of distributions. To overcome the limitations of existing density ratio-based methods, we propose a novel approach called the Pearson-like scaled-Bregman divergence-based (PLsBD) density ratio estimation method for change-point detection. Our theoretical studies derive an analytical expression for the Pearson-like scaled Bregman divergence using a mixture measure. We integrate the PLsBD with a kernel regression model and apply a random sampling strategy to identify change points in both synthetic data and real-world high-dimensional genomics data of Drosophila. Our PLsBD method demonstrates superior performance compared to many other change-point detection methods.
The purpose of anomaly detection is to identify special data points or patterns that significantly deviate from the expected or typical behavior of the majority of the data, and it has a wide range of applications across various domains. Most existing statistical and machine learning-based anomaly detection algorithms face challenges when applied to high-dimensional data. For instance, the unconstrained least-squares importance fitting (uLSIF) method, a state-of-the-art anomaly detection approach, encounters the unboundedness problem under certain conditions. In this study, we propose a scaled Bregman divergence-based anomaly detection algorithm using both least absolute deviation and least-squares loss for parameter learning. This new algorithm effectively addresses the unboundedness problem, making it particularly suitable for high-dimensional data. The proposed technique was evaluated on both synthetic and real-world high-dimensional time series datasets, demonstrating its effectiveness in detecting anomalies. Its performance was also compared to other density ratio estimation-based anomaly detection methods.
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.
Motivation: Missing data imputation remains a critical challenge in high-dimensional time-series data analysis, where traditional methods often struggle to capture complex nonlinear dependencies inherent in sequential data. Diffusion-based generative models have shown state-of-the-art performance by modeling the conditional distribution of missing values given observed data. However, these models typically rely on isotropic white noise during training, which can obscure important frequency-dependent correlations that are crucial for accurate imputation. Results: To address the limitations of conventional imputation methods, we propose a novel approach called time-varying blue noise-based conditional score-based diffusion model (tBN-CSDI). By modulating the noise schedule according to the frequency characteristics of the data, tBN-CSDI improves the recovery of subtle, high-frequency temporal patterns that are often overlooked by existing techniques. Experimental results on both healthcare and single-cell RNA-seq datasets show that tBN-CSDI consistently outperforms existing imputation methods, achieving over a 30% reduction in imputation error under high data sparsity. These findings underscore tBN-CSDI’s potential as a robust and effective solution for imputing sparse and noisy time-series data. We further discuss its practical applications in improving change-point detection and gene regulatory network inference, demonstrating its broader utility in biomedical and biological research. Availability and implementation: The computer code and data for the proposed method are available on GitHub: https://github.com/gbishop345/tBN-CSDI.
Inferring time-varying gene regulatory networks from time-series single-cell RNA sequencing (scRNA-seq) data remains a challenging task. The existing methods have notable limitations as most are either designed for reconstructing time-varying networks from bulk microarray data or constrained to inferring stationary networks from scRNA-seq data, failing to capture the dynamic regulatory changes at the single-cell level. Furthermore, scRNA-seq data present unique challenges, including sparsity, dropout events, and the need to account for heterogeneity across individual cells. These challenges complicate the accurate capture of gene regulatory network dynamics over time. In this work, we propose a novel f-divergence-based dynamic gene regulatory network inference method (f-DyGRN), which applies f-divergence to quantify the temporal variations in gene expression across individual single cells. Our approach integrates a first-order Granger causality model with various regularization techniques and partial correlation analysis to reconstruct gene regulatory networks from scRNA-seq data. To infer dynamic regulatory networks at different stages, we employ a moving window strategy, which allows for the capture of dynamic changes in gene interactions over time. We applied this method to analyze both simulated and real scRNA-seq data from THP-1 human myeloid monocytic leukemia cells, comparing its performance with the existing approaches. Our results demonstrate that f-DyGRN, when equipped with a suitable f-divergence measure, outperforms most of the existing methods in reconstructing dynamic regulatory networks from time-series scRNA-seq data.