SAX an innovative outlook to time-series representation

Aayush Kumar (~aayushkumarjvs)


Description:

Predictive analysis on Time Series Data Sets using SAX: The potential to do predictive machine learning on the data generated by connected sensors (temperature sensor, LIDAR etc) is a feature that is driving the spread of the Internet of Things. Predictive analysis of the time series data can be used to anticipate adverse events, enable early warning systems, improve results, reduce costs and enhance efficiency. An Anomaly (a.k.a outlier) is simply an unusual subsequence of the series. “Unusual” can be taken as “improbable”. Probability is not defined for the time series, Probability can be defined for Symbols. Mapping a time series to a symbol may allow us to assign a probability to the time series subsequence. This involves mapping the time series subsequence to a symbol in some symbol space. Key advantages of using SAX is that it yields an explainable model. Wherein the result of the model should not be underestimated. SAX is a good representation of working in raw data for most problems and representing time series data in the form of strings with a fixed-length size. SAX is a methodology for reducing a time series window to symbols. The technique was developed by Dr Eammonn Keogh et al UC Riverside in the early 2000s. It has drawn a great deal of attention in a world of time series analysis.

Prerequisites:

Machine Learning, Internet of Things, Predictive Analysis and Time Series Representation

Content URLs:

https://jmotif.github.io/sax-vsm_site/modules/algorithm/ https://r2s.hh.se/ReadingClub/2012-09-21/0-sax.pdf http://www.cse.cuhk.edu.hk/~adafu/Pub/icdm05time.pdf https://cs.gmu.edu/~jessica/sax.htm http://www.cs.ucr.edu/~eamonn/SAX.htm http://grammarviz2.github.io/grammarviz2_site/ https://github.com/johannfaouzi/pyts http://www.cs.ucr.edu/~eamonn/HOT%20SAX%20%20long-ver.pdf https://github.com/nphoff/saxpy http://alumni.cs.ucr.edu/~ratana/SSDBM05.pdf https://jmotif.github.io/sax-vsm_site/ https://www.cs.ucr.edu/~eamonn/discords/

Speaker Info:

Aayush Kumar is an undergrad at SRM IST, India, pursuing engineering in Information Technology. He loves to develop open-source software and sometimes surf Github more than facebook or twitter. Collaborating with people and making new things makes him happy! He is passionate about Artificial Intelligence, Machine Learning, Software Development, Open Source and all things related to technology. He loves challenges, competitions and creating an organized schedule for completing a task. Currently, He is working in the field of Data Analysis and Data Minning, He also loves to work on Computer Vision and Natural Language Understanding related problems.

Speaker Links:

Name: Aayush Kumar, Major: Information Technology, University: SRM Institute of Science and Technology, Github: https://github.com/aayushkumarjvs, Portfolio: https://aayushkumarjvs.ml, Slack: aayushkumarjvs, Email: aayushkumarjvs@gmail.com, Phone: (+91)755-016-7729, Postal Address: Flat no 401, Saptagiri Towers, Habsiguda, Hyderabad, Telengana State, India, 500 007. Timezone: Indian Standard Time (UTC +5:30)

Section: Data Science, Machine Learning and AI
Type: Poster
Target Audience: Intermediate
Last Updated: