Projects - Time Series Commons

Cosmo3DFlow: Wavelet Flow Matching for Spatial-to-Spectral Compression in Reconstructing the Early Universe

Cosmo3DFlow is a novel generative framework for reconstructing the early Universe from present-day observations. By combining 3D Discrete Wavelet Transform (DWT) flow matching with Temporal Homology (TH), it decouples high- and low-frequency cosmological structures for efficient spatial compression. Using N-body simulations, Cosmo3DFlow achieves 10x faster sampling than diffusion models, enabling initial conditions to be reconstructed in seconds rather than minutes. Read more

Large Language Models for Financial Aid in Financial Time-series Forecasting

This study utilizes Large Language Models (LLMs), such as pre-trained GPT-2, for financial time-series forecasting, addressing limited historical data and complex financial information. By benchmarking LLMs against state-of-the-art time-series models, the research highlights their superior predictive performance with minimal fine-tuning, offering valuable insights for financial decision-making. Read more

Interpreting Spatio-Temporal patterns with Self-Attention

This study uses the Temporal Fusion Transformer to forecast COVID-19 infections at the US county level, analyzing detailed temporal and spatial patterns from the self-attention to achieve superior prediction performance. By interpreting the model's learned patterns and using 2.5 years of socioeconomic and health data from 3,142 counties, this research provides valuable insights to aid effective public health decision-making. Read more

Interpreting Feature Interactions using Sensitivity Analysis

This study uses eight recent local interpretation methods on six Transformer-based time series models, comparing find the best predictor for COVID-19 cases using three years of data from 3,142 US counties. When also predicting the epidemic sensitivity to different age groups and compare the interpreted sensitivity with ground truth. The framework is also tested on other datasets for broader applicability. Read more

Windowed Temporal Saliency Analysis for Multi-Horizon Time Series Forecasting

Interpreting time series models is challenging due to the temporal dependencies between time steps and the changing significance of input features over time. This addresses these challenges by striving to provide clearer explanations of the feature interactions, and showcase the practical application of these interpretability techniques using real-world datasets and cutting-edge deep learning models. Read more

Interpretable AI for Spatio-Temporal Models

Cosmo3DFlow: Wavelet Flow Matching for Spatial-to-Spectral Compression in Reconstructing the Early Universe

Large Language Models for Financial Aid in Financial Time-series Forecasting

Interpreting Spatio-Temporal patterns with Self-Attention

Interpreting Feature Interactions using Sensitivity Analysis

Windowed Temporal Saliency Analysis for Multi-Horizon Time Series Forecasting

Videos

Software Packages Developed or Updated

Cosmo3DFlow: Wavelet Flow Matching for Early Universe Reconstruction

Large Language Models for Financial Aid

DS5110 Big Data Systems Fall 2024 AIForAstronomy Final Project

Interpreting County-Level COVID-19 Infections

Interpreting Time Series Transformer Models and Sensitivity Analysis of Population Age Groups to COVID-19 Infections

HySec-Flow: A Scalable and Secure Framework for Data Intensive Heterogeneous Computing

Surrogate Simulation for Earthquake Prediction