USAD - UnSupervised Anomaly Detection on multivariate time series

Last update: Jan 4, 2023

Related tags

Deep Learning usad

Overview

USAD - UnSupervised Anomaly Detection on multivariate time series

Scripts and utility programs for implementing the USAD architecture.

Implementation by: Francesco Galati.

Additional contributions: Julien Audibert, Maria A. Zuluaga.

How to cite

If you use this software, please cite the following paper as appropriate:

Audibert, J., Michiardi, P., Guyard, F., Marti, S., Zuluaga, M. A. (2020).
USAD : UnSupervised Anomaly Detection on multivariate time series.
Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, August 23-27, 2020

Requirements

PyTorch 1.6.0
CUDA 10.1 (to allow use of GPU, not compulsory)

Running the Software

All the python classes and functions strictly needed to implement the USAD architecture can be found in usad.py. An example of an application deployed with the SWaT dataset is included in USAD.ipynb.

Copyright and licensing

This software is released under the BSD-3 license. Please see the license file_ for details.

Publication

Audibert et al. USAD : UnSupervised Anomaly Detection on multivariate time series. 2020

Comments

Question about the Loss functions (L2 norm?)

Hello. May I ask about the loss functions inside the training_step? The paper states that ||x||2 denotes L2 Norm. As I understand, L2 norm is the "square root of sum of squares of elements inside x". However, the loss of the training step seems to be using MSE (mean((batch-w3)**2) etc). Was this a change from the paper or did I not understand L2 norm correctly? Thanks in advance!

opened by htakatsuka 2
execution class error

Thank you for your git project.

I run the code but the execution error is occured. Could you help me?

PS D:\deep_learning\anomaly_detection> python main.py --model USAD --action train C:\miniconda3\envs\yolov5\lib\site-packages\statsmodels\tools_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the fun ctions in the public API at pandas.testing instead. import pandas.util.testing as tm D:\deep_learning\anomaly_detection\dataPreprocessing.py:111: DtypeWarning: Columns (6,26,28,36,38,39,40,45,46,47) have mixed types.Specify dtype option on import or set low_memory=False. origin_normal, origin_attack, labels = self.SWAT_loadData(normal_data_path, attack_data_path) normal data.shape (495000, 51) #####data info######## number of anomaly : 54621 number of normal : 395298 ################
windows_normal.shape (494989, 12, 51) window_attack.shape (449908, 12, 51) w_size 612
z_size 480
self.train_loader <torch.utils.data.dataloader.DataLoader object at 0x00000156D30B19C8>
self.val_loader <torch.utils.data.dataloader.DataLoader object at 0x000001568A74FC08>
Traceback (most recent call last): File "main.py", line 154, in
executionObj.train(args.model) File "main.py", line 91, in train
history = self.model.training_all(self, n_epochs, self.train_loader, self.val_loader) File "D:\deep_learning\anomaly_detection\model.py", line 197, in training_all for epoch in range(epochs):
TypeError: 'execution' object cannot be interpreted as an integer

opened by leejunghan 0
Threshold Definition

Hi @manigalati , I am wondering how did you implement the threshold as AUC score? Isn't the AUC score is the kinda accuracy of the model. If you check the maximum value of y_pred, it is lower than the AUC score thus In the paper, they don't explain how did they set the threshold.

opened by ServetKocak 3
Fix memory usage during model validation

What was the problem: I couldn't train the model on my laptop with 4GB of VRAM.

What I have done: I disabled gradient calculation during model validation (gradients were stored in VRAM). Now, training full model as is (USAD.ipynb) takes about 1.8 GB of VRAM.

This PR should fix #6 .

opened by finloop 0
RuntimeError: CUDA out of memory.

I run this USAD.ipynb in jupyter-notebook.

train is ok. test is error. error message is below.

RuntimeError: CUDA out of memory. Tried to allocate 38.00 MiB (GPU 0; 10.76 GiB total capacity; 8.95 GiB already allocated; 17.06 MiB free; 9.61 GiB reserved in total by PyTorch)

=================================

How should I change the Parameter to solve this problem?

my gpu is GPU 0: GeForce RTX 2080 Ti, GPU 1: GeForce RTX 2080 Ti

Thank you.

opened by hongym7 2
Question about the parameters

Hello, can I ask why you set the batch_size=7919, and whether the z_size means latent space or not? Did you use the down-sampling rate mentioned in the article USAD? Thanks for your answer.

opened by JinchuanZ 27
Website visit 404

Hello, dear author， in usad.ipybn [12] WGet in file https://raw.githubusercontent.com/matthuisman/files.matthuisman.nz/master/gdrivedl . I can't find the website in https://raw.githubusercontent.com/matthuisman/files.matthuisman.nz/master/gdrivedl Website visit 404

opened by JunShengB 0

Owner

GitHub

Spectral Temporal Graph Neural Network (StemGNN in short) for Multivariate Time-series Forecasting

Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting This repository is the official implementation of Spectral Temporal Gr

306 Dec 29, 2022

Deep learning-based approach to discovering Granger causality networks in multivariate time series

Granger causality discovery for neural networks.

130 Jan 1, 2023

Code for the CIKM 2019 paper "DSANet: Dual Self-Attention Network for Multivariate Time Series Forecasting".

Dual Self-Attention Network for Multivariate Time Series Forecasting 20.10.26 Update: Due to the difficulty of installation and code maintenance cause

223 Dec 16, 2022

Library for implementing reservoir computing models (echo state networks) for multivariate time series classification and clustering.

Framework overview This library allows to quickly implement different architectures based on Reservoir Computing (the family of approaches popularized

249 Dec 21, 2022

The source code and data of the paper "Instance-wise Graph-based Framework for Multivariate Time Series Forecasting".

IGMTF The source code and data of the paper "Instance-wise Graph-based Framework for Multivariate Time Series Forecasting". Requirements The framework

24 Dec 5, 2022

Multivariate Time Series Forecasting with efficient Transformers. Code for the paper "Long-Range Transformers for Dynamic Spatiotemporal Forecasting."

Spacetimeformer Multivariate Forecasting This repository contains the code for the paper, "Long-Range Transformers for Dynamic Spatiotemporal Forecast

440 Jan 2, 2023

This repo contains the code required to train the multivariate time-series Transformer.

Multi-Variate Time-Series Transformer This repo contains the code required to train the multivariate time-series Transformer. Download the data The No

4 Nov 24, 2022

Official PyTorch code for WACV 2022 paper "CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows"

CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows WACV 2022 preprint:https://arxiv.org/abs/2107.1

156 Dec 28, 2022

Code for the paper "TadGAN: Time Series Anomaly Detection Using Generative Adversarial Networks"

TadGAN: Time Series Anomaly Detection Using Generative Adversarial Networks This is a Python3 / Pytorch implementation of TadGAN paper. The associated

92 Dec 3, 2022

Time-series-deep-learning - Developing Deep learning LSTM, BiLSTM models, and NeuralProphet for multi-step time-series forecasting of stock price.

Stock Price Prediction Using Deep Learning Univariate Time Series Predicting stock price using historical data of a company using Neural networks for

7 Nov 27, 2022

This is an unofficial implementation of the paper “Student-Teacher Feature Pyramid Matching for Unsupervised Anomaly Detection”.

32 Oct 26, 2022

USAD - UnSupervised Anomaly Detection on multivariate time series

Related tags

Overview

USAD - UnSupervised Anomaly Detection on multivariate time series

How to cite

Requirements

Running the Software

Copyright and licensing

Publication

Comments

Question about the Loss functions (L2 norm?)

execution class error

Threshold Definition

Fix memory usage during model validation

RuntimeError: CUDA out of memory.

Question about the parameters

Website visit 404

Owner

Spectral Temporal Graph Neural Network (StemGNN in short) for Multivariate Time-series Forecasting

Deep learning-based approach to discovering Granger causality networks in multivariate time series

Code for the CIKM 2019 paper "DSANet: Dual Self-Attention Network for Multivariate Time Series Forecasting".

Library for implementing reservoir computing models (echo state networks) for multivariate time series classification and clustering.

The source code and data of the paper "Instance-wise Graph-based Framework for Multivariate Time Series Forecasting".

Multivariate Time Series Forecasting with efficient Transformers. Code for the paper "Long-Range Transformers for Dynamic Spatiotemporal Forecasting."

This repo contains the code required to train the multivariate time-series Transformer.

Official PyTorch code for WACV 2022 paper "CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows"

Code for the paper "TadGAN: Time Series Anomaly Detection Using Generative Adversarial Networks"

Time-series-deep-learning - Developing Deep learning LSTM, BiLSTM models, and NeuralProphet for multi-step time-series forecasting of stock price.

This is an unofficial implementation of the paper “Student-Teacher Feature Pyramid Matching for Unsupervised Anomaly Detection”.

MvtecAD unsupervised Anomaly Detection

Demo project for real time anomaly detection using kafka and python

(JMLR'19) A Python Toolbox for Scalable Outlier Detection (Anomaly Detection)

Streaming Anomaly Detection Framework in Python (Outlier Detection for Streaming Data)

A Python Library for Graph Outlier Detection (Anomaly Detection)

Multivariate Boosted TRee

Real-world Anomaly Detection in Surveillance Videos- pytorch Re-implementation

Paper list of log-based anomaly detection