Mining Interpretable Spatio-temporal Logic Properties for Spatially Distributed Systems

The Internet-of-Things, complex sensor networks, multi-agent cyber-physical systems are all examples of spatially distributed systems that continuously evolve in time. Such systems generate huge amounts of spatio-temporal data, and system designers are often interested in analyzing and discovering structure within the data. There has been considerable interest in learning causal and logical properties of temporal data using logics such as Signal Temporal Logic (STL); however, there is limited work on discovering such relations on spatio-temporal data. We propose the first set of algorithms for unsupervised learning for spatio-temporal data. Our method does automatic feature extraction from the spatio-temporal data by projecting it onto the parameter space of a parametric spatio-temporal reach and escape logic (PSTREL). We propose an agglomerative hierarchical clustering technique that guarantees that each cluster satisfies a distinct STREL formula. We show that our method generates STREL formulas of bounded description complexity using a novel decision-tree approach which generalizes previous unsupervised learning techniques for Signal Temporal Logic. We demonstrate the effectiveness of our approach on case studies from diverse domains such as urban transportation, epidemiology, green infrastructure, and air quality monitoring.

[1]  Sanjit A. Seshia,et al.  Mining Requirements From Closed-Loop Control Models , 2015, IEEE Trans. Comput. Aided Des. Integr. Circuits Syst..

[2]  Y. Xing,et al.  The impact of PM2.5 on the human respiratory system. , 2016, Journal of thoracic disease.

[3]  Jyotirmoy V. Deshmukh,et al.  Interpretable classification of time-series data using efficient enumerative techniques , 2019, HSCC.

[4]  Dejan Nickovic,et al.  Parametric Identification of Temporal Properties , 2011, RV.

[5]  Eamonn J. Keogh,et al.  Clustering Time Series Using Unsupervised-Shapelets , 2012, 2012 IEEE 12th International Conference on Data Mining.

[6]  Vincenzo Ciancia,et al.  Qualitative and Quantitative Monitoring of Spatio-Temporal Properties with SSTL , 2017, Log. Methods Comput. Sci..

[7]  P. Rousseeuw Silhouettes: a graphical aid to the interpretation and validation of cluster analysis , 1987 .

[8]  Bernold Fiedler,et al.  Spatio-Temporal Dynamics of Reaction-Diffusion Patterns , 2003 .

[9]  H. Edelsbrunner,et al.  Efficient algorithms for agglomerative hierarchical clustering methods , 1984 .

[10]  Sanjit A. Seshia,et al.  Logical Clustering and Learning for Time-Series Data , 2016, CAV.

[11]  Jane Hillston,et al.  Probing the Performance of the Edinburgh Bike Sharing System using SSTL , 2020, SIGSIM-PADS.

[12]  Bhaskar Krishnamachari,et al.  COVID-19 Risk Estimation using a Time-varying SIR-model , 2020, COVID@SIGSPATIAL.

[13]  Ezio Bartocci,et al.  Monitoring mobile and spatially distributed cyber-physical systems , 2017, MEMOCODE.

[14]  Sanjit A. Seshia,et al.  Logical Clustering and Learning for Time-Series Data , 2016, 1612.07823.

[15]  Alberto L. Sangiovanni-Vincentelli,et al.  Time Series Learning using Monotonic Logical Properties , 2018, RV.

[16]  Dejan Nickovic,et al.  Monitoring Temporal Properties of Continuous Signals , 2004, FORMATS/FTRTFT.

[17]  Marc Rußwurm,et al.  Tslearn, A Machine Learning Toolkit for Time Series Data , 2020, J. Mach. Learn. Res..

[18]  Jyotirmoy V. Deshmukh,et al.  Mining Environment Assumptions for Cyber-Physical System Models , 2020, 2020 ACM/IEEE 11th International Conference on Cyber-Physical Systems (ICCPS).

[19]  Ezio Bartocci,et al.  MoonLight: a lightweight tool for monitoring spatio-temporal properties , 2020, International Journal on Software Tools for Technology Transfer.