Download as PDF, PPTX
![Clustering subsequence is meaningless [1]. Eamonn Keogh, Jessica Lin, Clustering of Time Series Subsequences is Meaningless: Implications for Previous and Future Research](https://image.slidesharecdn.com/2-190718140733/75/Autoencoder-Forest-for-Anomaly-Detection-from-IoT-Time-Series-12-2048.jpg)
Uploaded byYiqun Hu
Autoencoder Forest for Anomaly Detection from IoT Time Series
The document discusses the use of autoencoder forests for anomaly detection in IoT time series data, highlighting the inefficiencies of manual monitoring and traditional rule-based methods. It explains the structure and training process of autoencoding neural networks, particularly in handling time series analysis and anomaly scoring. Experiment results are presented, demonstrating its application in various monitoring scenarios, such as cooling tower temperatures and energy consumption.
In this document
Powered by AI
Introduction to the Autoencoder Forest for IoT Time Series anomaly detection and the agenda of the presentation.
Explains condition monitoring and the need for anomaly detection, highlighting issues with manual monitoring and rule-based methods.
Describes Autoencoders, a neural network type used for anomaly detection, including its online detection and reconstruction error concepts.
Addresses the limitations of single autoencoders and introduces the Autoencoder Forest concept, emphasizing its clustering advantages.
Details the architecture and training approach for Autoencoder Forest, with emphasis on parallel training and early stopping mechanisms.
Describes the systematic workflow for conducting anomaly detection using Autoencoder Forest, from preprocessing to anomaly scoring.
Explains methods for automatic periodic pattern analysis in time series, including autocorrelation calculations.
Discusses strategies for managing missing data in time series analysis without imputation for small gaps.
Outlines the method for extracting and scoring sequences to determine anomalies using the trained Autoencoder Forest.
Presents experimental results and types of anomalies detected across different parameters in various systems.
Introduces a unified platform for time series data analysis that incorporates built-in AI capabilities.
Autoencoder Forest for Anomaly Detection from IoT Time Series
- 1. <Title> Autoencoder Forest forAnomaly Detection from IoT Time Series Yiqun Hu, SP Group Data Council Singapore 2019
- 2. Agenda • Condition monitoring& anomaly detection • Autoencoder for anomaly detection • Autoencoder Forest • End-to-end workflow • Experiment results
- 3. Conditional monitoring &Anomaly Detection
- 4.
- 5. • Manual monitoring –Huge human effort – Boring task with low quality • Rule-based method – Cannot differentiate different environment – Cannot adapt to different condition of the equipment • Data-driven method – Model the common behavior of the equipment Time-series anomaly detection
- 6. Autoencoder for AnomalyDetection
- 7. Autoencoder • What isautoencoder – A encoder-decoder type of neural network architecture that is used for self-learning from unlabeled data • The idea of autoencoder – Learn how to compress data into a concise representation to allow for the reconstruction with minimum error • Different variants of autoencoder – Variational Autoencoder – LSTM Autoencoder – Etc. Autoencoder Neural Network
- 8. Autoencoder for anomalydetection Online Detection Anomaly score Offline Training Reconstruction errors
- 9.
- 10. A key challengeof autoencoder Single Autoencoder
- 11. The idea ofautoencoder forest x x x xx x x x x o o o o o o o o o + ++ + + + +
- 12. Clustering subsequence ismeaningless [1]. Eamonn Keogh, Jessica Lin, Clustering of Time Series Subsequences is Meaningless: Implications for Previous and Future Research
- 13. Autoencoder forest basedon time 0:00 1:00 1:30 22:00 23:30
- 14. Training autoencoder forest InputLayer Encoder layer 1 (window_size, 1) (window_size/2, 1) Encoder layer 2 (window_size/4, 1) Decoder layer 1 (window_size/2, 1) Decoder Layer 2 (window_size, 1) • Structure is fixed for every autoencoder. (try to make it as generic as possible) • Each autoencoder within forest is independent. So the training is naturally parallelizable • Using early stopping mechanism, the training of individual autoencoder can be stopped at similar accuracy.
- 15. Autoencoder Forest Single AutoencoderAutoencoder Forest
- 16.
- 17. Automatic end-to-end workflow Timeseries analysis Train Data Preprocessing Train Window Extraction Autoencoder Forest Training Test Data Preprocessing Test Window Extraction Anomaly scoring Training Anomaly detection
- 18. Periodic pattern analysis •Automatic determine the repeating period in time series – Calculate autocorrelations of different lags – Find the strong local maximum of autocorrelation – Calculate the interval of any two local maximum – Find the mode of intervala
- 19. Missing data handling 3:053:10 3:15 3:20 … … 16:15 16:21 16:24 16:30 … … Misalignment Missing 3:05 3:10 3:15 3:20 … … 16:15 (16:20 – 16:40) 16:45 … … ? ? ? • No need to impute • If missing gap is small, impute with neighbouring points; • If missing gap is large, impute with the same time of other periods;
- 20. Anomaly scoring Extract thesequence window end at time t ...... Median profile Corresponding autoencoder reconstruct the sequence window at time t Compute reconstruction error as anomaly score Learned autoencoder forest
- 21.
- 22. Cooling tower –return water temperature
- 23. Chiller – chilledwater return temperature
- 24. Smart meter –half hour consumption 2018-12-03 22:00:00 Normal data 2018-09-27 14:30:00 2018-10-06 22:30:00 2018-09-07 15:30:00 Top 3 Detected Anomaly
- 26. A common platformfor time series data, with built-in AI capabilities
- 27.