A Fresh Perspective: Learning to Sparsify for Detection in Massive Noisy Sensor Networks IPSN 4/9/2013 Matthew Faulkner Annie Liu Andreas Krause.

Slides:



Advertisements
Similar presentations
Bayesian Belief Propagation
Advertisements

theory and practice collide The Next Big One: Detecting Earthquakes and Other Rare Events from Community Sensors Matthew Faulkner,
A Hierarchical Multiple Target Tracking Algorithm for Sensor Networks Songhwai Oh and Shankar Sastry EECS, Berkeley Nest Retreat, Jan
EE645: Independent Component Analysis
IODetector: A Generic Service for Indoor Outdoor Detection Pengfei Zhou†, Yuanqing Zheng†, Zhenjiang Li†, Mo Li†, and Guobin Shen‡ †Nanyang Technological.
EE462 MLCV Lecture 5-6 Object Detection – Boosting Tae-Kyun Kim.
Cost-effective Outbreak Detection in Networks Jure Leskovec, Andreas Krause, Carlos Guestrin, Christos Faloutsos, Jeanne VanBriesen, Natalie Glance.
Online Distributed Sensor Selection Daniel Golovin, Matthew Faulkner, Andreas Krause theory and practice collide 1.
Rapid Detection of Rare Geospatial Events: Earthquake Warning Applications A Review by Zahid Mian WPI CS525D September 10, 2012.
Large wireless autonomic networks Sensor networks Philippe Jacquet.
Analog and digital data Skills: none IT concepts: analog to digital conversion, digital to analog conversion, sample rate, sample size, quality-file size.
Real Time Motion Capture Using a Single Time-Of-Flight Camera
Data centres and observablesModern Seismology – Data processing and inversion 1 Data in seismology: networks, instruments, current problems  Seismic networks,
1/24 Passive Interference Measurement in Wireless Sensor Networks Shucheng Liu 1,2, Guoliang Xing 3, Hongwei Zhang 4, Jianping Wang 2, Jun Huang 3, Mo.
Distributed Regression: an Efficient Framework for Modeling Sensor Network Data Carlos Guestrin Peter Bodik Romain Thibaux Mark Paskin Samuel Madden.
Watchdog Confident Event Detection in Heterogeneous Sensor Networks Matthew Keally 1, Gang Zhou 1, Guoliang Xing 2 1 College of William and Mary, 2 Michigan.
INFERRING NETWORKS OF DIFFUSION AND INFLUENCE Presented by Alicia Frame Paper by Manuel Gomez-Rodriguez, Jure Leskovec, and Andreas Kraus.
Chess Review May 11, 2005 Berkeley, CA Tracking Multiple Objects using Sensor Networks and Camera Networks Songhwai Oh EECS, UC Berkeley
Scalable Training of Mixture Models via Coresets Daniel Feldman Matthew Faulkner Andreas Krause MIT.
Model-driven Data Acquisition in Sensor Networks Amol Deshpande 1,4 Carlos Guestrin 4,2 Sam Madden 4,3 Joe Hellerstein 1,4 Wei Hong 4 1 UC Berkeley 2 Carnegie.
HELSINKI UNIVERSITY OF TECHNOLOGY LABORATORY OF COMPUTER AND INFORMATION SCIENCE NEURAL NETWORKS RESEACH CENTRE Variability of Independent Components.
Hierarchical Trust Management for Wireless Sensor Networks and Its Application to Trust-Based Routing Fenye Bao, Ing-Ray Chen, Moonjeong Chang Presented.
Sensor Positioning in Wireless Ad-hoc Sensor Networks Using Multidimensional Scaling Xiang Ji and Hongyuan Zha Dept. of Computer Science and Engineering,
Infrasound detector for Apatity group Asming V.E., Kola Regional Seismological Center, Apatity, Russia.
PRESENTED BY: SAURAV SINGH.  INTRODUCTION  WINS SYSTEM ARCHITECTURE  WINS NODE ARCHITECTURE  WINS MICRO SENSORS  DISTRIBUTED SENSOR AT BORDER  WINS.
IE 585 Introduction to Neural Networks. 2 Modeling Continuum Unarticulated Wisdom Articulated Qualitative Models Theoretic (First Principles) Models Empirical.
1 1 Stanford University 2 MPI for Biological Cybernetics 3 California Institute of Technology Inferring Networks of Diffusion and Influence Manuel Gomez.
1 1 Stanford University 2 MPI for Biological Cybernetics 3 California Institute of Technology Inferring Networks of Diffusion and Influence Manuel Gomez.
REVISED CONTEXTUAL LRT FOR VOICE ACTIVITY DETECTION Javier Ram’ırez, Jos’e C. Segura and J.M. G’orriz Dept. of Signal Theory Networking and Communications.
Influence Maximization in Dynamic Social Networks Honglei Zhuang, Yihan Sun, Jie Tang, Jialin Zhang, Xiaoming Sun.
23/10/2015 E.R.Edwards 23/10/2015 Staffordshire University School of Computing Introduction to Android Sensors.
Beginning search for deep moonquakes locations on the lunar far side T. Sonnemann 1, 2, M. Knapmeyer 1 and J. Oberst 1 1 Institute of Planetary Research,
28 February, 2003University of Glasgow1 Cluster Variation Method and Probabilistic Image Processing -- Loopy Belief Propagation -- Kazuyuki Tanaka Graduate.
Seismological Perspectives on Broadband Tilt, Strain, and Rotation Measurement Charles Langston Center for Earthquake Research and Information University.
A Passive Approach to Sensor Network Localization Rahul Biswas and Sebastian Thrun International Conference on Intelligent Robots and Systems 2004 Presented.
Using Information Technology to Reduce Traffic Jam in a Highly Traffic Congested City Sayed Ahmed and Rasit Eskicioglu We propose a cost effective and.
Junfeng Xu, Keqiu Li, and Geyong Min IEEE Globecom 2010 Speak: Huei-Rung, Tsai Layered Multi-path Power Control in Underwater Sensor Networks.
High-integrity Sensor Networks Mani Srivastava UCLA.
Yanlei Diao, University of Massachusetts Amherst Future Directions in Sensor Data Management Yanlei Diao University of Massachusetts, Amherst.
Radiation Detection and Measurement, JU, First Semester, (Saed Dababneh). 1 Counting Statistics and Error Prediction Poisson Distribution ( p.
Srinivas Cheekati( ) Instructor: Dr. Dong-Chul Kim
Syed Hassan Ahmed Syed Hassan Ahmed, Safdar H. Bouk, Nadeem Javaid, and Iwao Sasase RIU Islamabad. IMNIC’12, RIU Islamabad.
1 Lecture 6 Neural Network Training. 2 Neural Network Training Network training is basic to establishing the functional relationship between the inputs.
Radiation Detection and Measurement, JU, 1st Semester, (Saed Dababneh). 1 Radioactive decay is a random process. Fluctuations. Characterization.
Sensors in android. App being more applicable Keeping track of your heart beat while jogging. Pointing the phone camera towards the night sky to know.
A novel methodology for identification of inhomogeneities in climate time series Andrés Farall 1, Jean-Phillipe Boulanger 1, Liliana Orellana 2 1 CLARIS.
The ambient light sensor in a smart phone is what measures how bright the light is. It’s the phones software that adjusts the brightness in the display.
Cameron Rowe.  Introduction  Purpose  Implementation  Simple Example Problem  Extended Kalman Filters  Conclusion  Real World Examples.
Sensors For Mobile Phones  Ambient Light Sensor  Proximity Sensor  GPS Receiver Sensor  Gyroscope Sensor  Barometer Sensor  Accelerometer Sensor.
Combined Human, Antenna Orientation in Elevation Direction and Ground Effect on RSSI in Wireless Sensor Networks Syed Hassan Ahmed, Safdar H. Bouk, Nadeem.
Container Database Management Zheng Liu, Sheng Liu CSE 534:Advanced computer networks.
1 1 Stanford University 2 MPI for Biological Cybernetics 3 California Institute of Technology Inferring Networks of Diffusion and Influence Manuel Gomez.
Perfect recall: Every decision node observes all earlier decision nodes and their parents (along a “temporal” order) Sum-max-sum rule (dynamical programming):
BORDER SECURITY USING WIRELESS INTEGRATED NETWORK SENSORS (WINS) By B.S.Indrani (07841A0406) Aurora’s Technological and Research Institute.
Border Security Using Wireless Integrated Network Sensors
IShake System: Earthquake Detection with Smartphones Presenter: Jize Zhang Da Huo Original Paper:Reilly, Jack, et al. "Mobile phones as seismologic sensors:
Make an information leaflet about what the sensors do in a Smart Phone for people over 65 years of age. You can use PowerPoint, Word or Publisher.
Smart City A Public-Private Partnership. Uses communication networks, wireless sensor technology and intelligent data management to make decisions in.
Sensor Networks © M Jamshidi.
Inferring Networks of Diffusion and Influence
Border security using Wireless Integrated Network Sensors(WINS)
Near-optimal Observation Selection using Submodular Functions
MART: Music Assisted Running Trainer
Google's 1Gbps Network Introduction Little about Google
Faulkner, Matthew, Michael Olson, Rishi Chandy, Jonathan Krause, K
Cost-effective Outbreak Detection in Networks
An Infant Facial Expression Recognition System Based on Moment Feature Extraction C. Y. Fang, H. W. Lin, S. W. Chen Department of Computer Science and.
Origin of Universe - Big Bang
Chapter 5 – Distributed Elements
Information Sciences and Systems Lab
Presentation transcript:

A Fresh Perspective: Learning to Sparsify for Detection in Massive Noisy Sensor Networks IPSN 4/9/2013 Matthew Faulkner Annie Liu Andreas Krause

Community Sensors More than 1 Billion smart devices provide powerful internet-connected sensor packages. Video Sound GPS Acceleration Rotation Temperature Magnetic Field Light Humidity Proximity

Dense, City-wide Networks Signal Hill Seismic Survey 5000 Seismometers What could dense networks measure?

Dense, City-wide Networks What could dense networks measure? Signal Hill Seismic Survey 5000 Sesimometers

Long Beach Seismic Network

Caltech Community Seismic Network Detecting and Measuring quakes with community sensors 16-bit USB Accelerometer CSN-Droid Android App

Scaling with Decentralized Detection Quake? 5000 Long Beach: 250 GB/day 300K LA: 15 TB/day

Scaling with Decentralized Detection Optimal decentralized tests Hypothesis testing [Tsitsiklis ‘88] Local Detection Quake? But strong assumptions…

9 ‘Weak’ Signals in Massive Networks No pick Pick

10 ‘Weak’ Signals in Massive Networks No pick Pick

11 ‘Weak’ Signals in Massive Networks No pick Pick

12 ‘Weak’ Signals in Massive Networks No pick Pick

Trading Quantity for Quality? Detecting arbitrary weak signals requires diminishing noise

“Sparsifiable” Events

A Basis from Clustering Hierarchical clustering defines an orthonormal basis Haar Wavelet Basis

Latent Tree Model Hierarchical dependencies can produce sparsifiable signals.

Latent Tree Model Hierarchical dependencies can produce sparsifiable signals.

From Sparsification to Detection Applying the basis to observed data gives a detection rule Lots of noisy sensors can be reliable!

Learning a Sparsifying Basis Given real data, can we learn a sparsifying basis? ICA [Hyvärinen & Oja ‘00] Efficient, but assumes noise-free observations X Continuous, smooth

Learning a Sparsifying Basis Given real data, can we learn a sparsifying basis? SLSA [Chen 2011] Learns the basis from noisy data

Synthetic Experiments Event signals generated from Singh’s Latent Tree Model Gaussian noise Binary noise Learned bases (ICA, SLSA) outperform baseline average and wavelet basis Noise VarianceBinary Error Rate

Outbreaks on Gnutella P2P 1769 High-degree nodes in the Gnutella P2P network. snap.stanford.edu 40,000 simulated cascades. AUC(0.05) Learned bases (SLSA, ICA) outperform scan statistics Binary noise rate

Japan Seismic Network quakes recorded after the 2011 Tohoku M9.0 quake 721 Hi-net seismometers AUC(0.001) – small tolerance to false positive Binary noise rate

Japan Seismic Network Learned basis elements capture wave propagation AUC(0.001) – small tolerance to false positive Binary noise rate

Long Beach Sesimic Network 1,000 sensors Five M2.5 - M3.4 quakes

Long Beach Seismic Network 2000 simulated quakes provide training data Learned bases (SLSA, ICA) outperform wavelet basis and scan statistics

Caltech Community Seismic Network 128 sensors Four M3.2 – M5.4 quakes

Caltech Community Seismic Network Trained on 1,000 simulated quakes Learned bases (SLSA, ICA) detect quakes up to 8 seconds faster

Conclusions Theoretical guarantees about decentralized detection of sparsifiable events Framework for learning sparsifying bases from simulations or sensor measurements Strong experimental performance on 3 seismic networks, and simulated epidemics in P2P networks Real-time event detection in massive, noisy community sensor networks