WenjieDu/SAITS
The official PyTorch implementation of the paper "SAITS: Self-Attention-based Imputation for Time Series". A fast and state-of-the-art (SOTA) deep-learning neural network model for efficient time-series imputation (impute multivariate incomplete time series containing NaN missing data/values with machine learning). https://arxiv.org/abs/2202.08516
This tool helps scientists, traders, or operations engineers analyze complex time-series data, even when some measurements are missing. It takes incomplete historical data, like sensor readings or stock prices with gaps, and intelligently fills in the missing values. The output is a complete, imputed dataset ready for further analysis, forecasting, or decision-making.
498 stars. No commits in the last 6 months.
Use this if you have multivariate time-series datasets with missing values (NaNs) and need a robust, state-of-the-art method to accurately fill those gaps using deep learning.
Not ideal if your data is not time-series based, you need a simple, interpretable imputation method, or you lack the computational resources for deep learning models.
Stars
498
Forks
70
Language
Python
License
MIT
Category
Last pushed
Oct 01, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/WenjieDu/SAITS"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
sktime/sktime
A unified framework for machine learning with time series
aeon-toolkit/aeon
A toolkit for time series machine learning and deep learning
Nixtla/neuralforecast
Scalable and user friendly neural :brain: forecasting algorithms.
tslearn-team/tslearn
The machine learning toolkit for time series analysis in Python
Nixtla/mlforecast
Scalable machine 🤖 learning for time series forecasting.