CrabNet
SAITS
CrabNet | SAITS | |
---|---|---|
1 | 9 | |
96 | 361 | |
- | 3.3% | |
3.7 | 6.2 | |
almost 2 years ago | 5 months ago | |
Python | Python | |
MIT License | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
CrabNet
-
Artificial intelligence can revolutionise science
I don't know. As for "literature-based discovery," this project/paper sounded like a pretty big deal when it came out a few years ago: https://github.com/materialsintelligence/mat2vec . And I see this thing came out more recently: https://github.com/anthony-wang/CrabNet .
Of course not all fields lend themselves as well to this as does materials science.
SAITS
-
SAITS: Self-Attention-based Imputation for Time Series. Expert Systems with Applications, 219:119619, 2023.
The full paper is available on arXiv: https://arxiv.org/abs/2202.08516. The code on GitHub: https://github.com/WenjieDu/SAITS/. If your research lies in time-series modeling, you may also be interested in the work PyPOTS: a Python toolbox for data mining on Partially-Observed Time Series. Its full paper is available on arXiv as well https://arxiv.org/abs/2305.18811, which has been peer-reviewed and accepted by the 9th SIGKDD international workshop Mining and Learning from Time Series (MiLeTS'23).
-
[R] SAITS: Self-Attention-based Imputation for Time Series. Expert Systems with Applications, 219:119619, 2023.
Missing values in time series collected from the real world are common to see and very pesky. A new state-of-the-art and fast neural network called "SAITS“ is proposed to help impute missing data in partially-observed multivariate time series. The paper has been peer-reviewed and published in the journal Expert Systems with Applications (DOI link). The full paper is available on arXiv at this URL. The code is open source on GitHub https://github.com/WenjieDu/SAITS/.
-
Missing values in time series collected from the real world are common to see and very pesky. A new state-of-the-art and fast neural network called SAITS is proposed to impute missing data in partially-observed multivariate time series. The code is open source on GitHub.
The code on GitHub: https://github.com/WenjieDu/SAITS/.
- SAITS: NEW Data - star count:118.0
What are some alternatives?
query-selector - LONG-TERM SERIES FORECASTING WITH QUERYSELECTOR – EFFICIENT MODEL OF SPARSEATTENTION
ViTGAN - A PyTorch implementation of ViTGAN based on paper ViTGAN: Training GANs with Vision Transformers.
ProteinStructurePrediction - Protein structure prediction is the task of predicting the 3-dimensional structure (shape) of a protein given its amino acid sequence and any available supporting information. In this section, we will Install and inspect sidechainnet, a dataset with tools for predicting and inspecting protein structures, complete two simplified implementations of Attention based Networks for predicting protein angles from amino acid sequences, and visualize our predictions along the way.
Informer2020 - The GitHub repository for the paper "Informer" accepted by AAAI 2021.
mat2vec - Supplementary Materials for Tshitoyan et al. "Unsupervised word embeddings capture latent knowledge from materials science literature", Nature (2019).
flashattention2-custom-mask - Triton implementation of FlashAttention2 that adds Custom Masks.
Regressio - A python library for univariate regression, interpolation, and smoothing.
GAT - Graph Attention Networks (https://arxiv.org/abs/1710.10903)
SSSD - Repository for the paper: 'Diffusion-based Time Series Imputation and Forecasting with Structured State Space Models'
Invariant-Attention - An implementation of Invariant Point Attention from Alphafold 2
PyPOTS - A Python toolkit/library for reality-centric machine/deep learning and data mining on partially-observed time series, including SOTA neural network models for scientific analysis tasks of imputation/classification/clustering/forecasting/anomaly detection/cleaning on incomplete industrial (irregularly-sampled) multivariate TS with NaN missing values