Big Data Challenges in Large IP Networks

Size: px
Start display at page:

Download "Big Data Challenges in Large IP Networks"

Transcription

1 Big Data Challenges in Large IP Networks Feature Extraction & Predictive Alarms for network management Wednesday 28 th Feb 2018 Dave Yearling British Telecommunications plc 2017

2 What we will cover Making data connections for our Customers Each Router 10 line cards Each Line card maps 40 ports Performance monitored (every 15minutes) via 12 metrics 30m Premises 5k Exchanges 1k Edge Routers Sensors on Chassis, CPU.. 5m time series 0.5 billion obs/day The challenge - predicting failures in routers for in time recovery Cannot predict everything (e.g. failure due to power outage, third party damage, ) But perhaps we can Summarise metrics to hourly Extract Features here we focus on one technique Cortical Learning Algorithm (CLA) Predict line card failures & generate informative alarms Solutions need to be deployed on Big Data platform Large numbers of heterogeneous times series Scaling algorithms for real time streaming data, 1 observation at a time More than one mode of failure, failures are rare Predictions need to have low false positive rate

3 Vital for our customers What we do now and our aim IP Traffic and load set to grow massively 4G (and future 5G) relies on 21 CN QoS Customer tolerance of poor service will decrease & 1 Outage can affect many customers Monitor and apply threshold alarms Some straightforward & essential e.g. Temperature Others complex & poorly understood Many false alarms, nothing recorded for near miss Many do not give Next Best Action' Fast Reactive Logic Provides swift reaction but After the event Dynamically monitor customers on demand testing Identifies common network elements Provides a Next Best Action Timely Predictive Logic Currently NO Predictive capability for raw metrics Distil, model and apply On line 1 record at a time Provides a probabilistic view of all network elements augmenting alarms Storing multiple models & data e.g. Card Failure, Degraded Service,

4 The road map for predicting card failures Resource A Metrics For each router, card, port, A B C D E F G H Resource H Metrics but there are Challenges 1. PROCESS For Every Resource/Metric Clean and snap to regular intervals 1. PROCESS Some detectors require windows of time for each point of measurement 2. FEATURE EXTRACTION Run Anomaly Detector Suite Data reduction Produce patterns for Diagnostics 3. PREDICTION Data Volume greatly reduced Only process anomalies Continuous M/Learning Desirable 2. FEATURE EXTRACTION Heterogeneous time series so we are planning an ensemble Running at real time is difficult Business lacks highly specific expertise - PhDs with Lancaster Multivariate structure Missing values 3. PREDICTION Input summary is heuristic but done once for all predictive models Rare events means we have to do sandpit builds Rate of change of probabilities may provide more information Current Ensemble include Cortical Learning Algorithm PELT/Binary Segmentation Hybrid Extreme Stud/Dev HOT SAX 4. APPLY Record Probabilities for each outcome Flag Warnings & visualise probabilities (in space & time) 4. APPLY Need to capture feedback from operations in order to tune Consuming this into the Operations is non trivial

5 Wouldn t it be great if we had an algorithm that behaves just like an operation analyst We need Feature Extraction to flag strangeness from heterogeneous time series METRIC I METRIC II Resource A Resource A Must cope with Periodicity Huge Transients Developing Anomalies Resource B Resource B And detect changes in Baseline Mean Variance Resource C Resource C And have low false alert rate Point of Failure Point of Failure Strangeness here are outliers : an observation (or a subset of observations) which appears to be inconsistent with the remainder of that set of data* * Barnett & Lewis, 1994

6 Cortical Learning Algorithm (CLA) Follows principles of the human neocortex Higher functions accomplished mostly within uniform crinkled surface of neocortex Different functions (vision, speech..) associated with lobes but very importantly Functions operate in the same way & on the same data signals We recognise/solve/produce behaviour by storing & recalling memories as SEQUENCES These SEQUENCES are stored in an INVARIANT and NOT Exhaustive way Continuously compares input with its own predictions on what to expect next In essence CLA is a self learning anomaly detection platform handles streaming data 1 observation at a time requires to store model not time series big advantage with many tuning parameters Please see Numenta Platform for Intelligent Computer (NuPIC) Numenta.org for much more detail on CLA

7 Cortical Learning Algorithm Anomaly Detector in a nutshell Inbound Data Stream ENCODER ENCODE observation at time t into binary vector All Columns relate to some region of input SPARSE DISTRIBUTED REPRESENTATION Cortical Columns representing the Layers Use approach 32 cells per column e.g of 2k bits only 2% (i.e. 40 bits) can be set to 1, and each of these have semantic meaning SPATIAL POOLER TEMPORAL POOLER ANOMALY CALCULATION RECOGNISE input, similar vectors - similar columns are active 1. Calculate Overlap for each column (a degree of how well they map) 2. Activate top 2% of Overlap columns, subject to Inhibition 3. Adjust perm for all synapses on Activated Columns - learning 1. Each synapse which has a degree of permanence perm (0,1) 2. Allows synapse to fire when perm > τ (some threshold) PREDICT next input vector based on the current/previous states For cells in each Active column : 1. Switch all PREDICTIVE cells to ACTIVE, if NONE make ALL cells to ACTIVE 2. Inter column connections are formed/reinforced between all ACTIVE cells in ACTIVE columns to the WINNING CELLS at previous time step 3. Set a cell as most predictive WINNING Cell in each ACTIVE COLUMN (according to a set of rules) ANOMALY based on the correctness of previous prediction(s) (i) % of spatial pooler at t incorrectly predicted via temporal pooler at t-1 (ii) Calculate RECENT Moving Average μ t compare with LONG TERM M/Average μ t Compare tail probability μ t μ t ~N(0,1) σ t 2,048 columns all map back to fixed area on the SDR (all cells in column inherit this) cells have states INACTIVE PREDICTIVE ACTIVE cells connect to cells in other columns These are connected & disconnected by using perm with threshold

8 Cortical Learning Algorithm Experiment & example time series CARD OUTAGE Thought experiment Measure all series at 5 days prior to any Card outage (for Targets - YES) and 15 th Sept (for Controls - NO) Cohorts of circuits YES Any Card Outage in 5 days, 158 observations NO No Outage of any kind, 648 observations Simple Heuristic At each point we sum all weighted anomalies detected in the past 15 days (s.t. anomalies on the t -15 th day are worth 10% compared to t)

9 Cortical Learning Algorithm Experiment & example time series CARD OUTAGE Thought experiment Measure all series at 5 days prior to any Card outage (for Targets - YES) and 15 th Sept (for Controls - NO) Cohorts of circuits YES Any Card Outage in 5 days, 158 observations NO No Outage of any kind, 648 observations Simple Heuristic At each point we sum all weighted anomalies detected in the past 15 days (s.t. anomalies on the t -15 th day are worth 10% compared to t)

10 Cortical Learning Algorithm Experiment Likelihood Threshold 0.95, 15 day exponential weighted window Comparing all circuits within both YES and NO Card Failure Cohorts xperi Day 0 = Day 1 = Day 15 = 0. 1

11 In Summary Recap and Current Focus The Techniques The Business Current Focus CLA..has promise Deals with streaming but Not currently in real time Anomaly flagging can be inconsistent Parameter tuning exploration When to switch off/on learning Other Ensemble methods Typically much faster, but: generates lots of anomalies relies on window processing Require more pre-processing Missing Values - Robust ARIMA & Kalman Filter Multivariate structure not yet dealt with Monitoring these data streams and suggesting intervention remains daunting Outcomes we wish to predict are: not always cleanly defined (e.g. Error occurred somewhere within the router) very rare occurrences Moving the models & detectors to the data difficult multi-tenanted Hadoop environment operational systems security & long release cycles Continued work on Feature Extraction - augmenting with current alarm output Remove redundancy non informative or collinear metrics Collaboration with operational analysts & software engineers Provide Feature Extraction results for root cause analysis

Spatial Pooler Algorithm Implementation and Pseudocode

Spatial Pooler Algorithm Implementation and Pseudocode Chapter Revision History The table notes major changes between revisions. Minor changes such as small clarifications or formatting changes are not noted. Version Date Changes Principal Author(s) 0.4 Initial

More information

Artificial Intelligence Introduction Handwriting Recognition Kadir Eren Unal ( ), Jakob Heyder ( )

Artificial Intelligence Introduction Handwriting Recognition Kadir Eren Unal ( ), Jakob Heyder ( ) Structure: 1. Introduction 2. Problem 3. Neural network approach a. Architecture b. Phases of CNN c. Results 4. HTM approach a. Architecture b. Setup c. Results 5. Conclusion 1.) Introduction Artificial

More information

Problem Code: #ISR13. College Code :

Problem Code: #ISR13. College Code : Ministry Category : Indian Space Research Organisation (ISRO) Problem statement : Detect security breaches or anomalies in Network traffic using Network log analysis Team Leader Name : Vipin George Problem

More information

Best Practices for Alert Tuning. This white paper will provide best practices for alert tuning to ensure two related outcomes:

Best Practices for Alert Tuning. This white paper will provide best practices for alert tuning to ensure two related outcomes: This white paper will provide best practices for alert tuning to ensure two related outcomes: 1. Monitoring is in place to catch critical conditions and alert the right people 2. Noise is reduced and people

More information

Video AI Alerts An Artificial Intelligence-Based Approach to Anomaly Detection and Root Cause Analysis for OTT Video Publishers

Video AI Alerts An Artificial Intelligence-Based Approach to Anomaly Detection and Root Cause Analysis for OTT Video Publishers Video AI Alerts An Artificial Intelligence-Based Approach to Anomaly Detection and Root Cause Analysis for OTT Video Publishers Live and on-demand programming delivered by over-the-top (OTT) will soon

More information

5 STEPS for Turning Data into Actionable Insights

5 STEPS for Turning Data into Actionable Insights 5 STEPS for Turning Data into Actionable Insights Today, nearly every piece of equipment in a data center has at least one sensor attached to it. In fact, American Society of Heating, Refrigerating and

More information

Optimizing Parallel Access to the BaBar Database System Using CORBA Servers

Optimizing Parallel Access to the BaBar Database System Using CORBA Servers SLAC-PUB-9176 September 2001 Optimizing Parallel Access to the BaBar Database System Using CORBA Servers Jacek Becla 1, Igor Gaponenko 2 1 Stanford Linear Accelerator Center Stanford University, Stanford,

More information

DATA MINING AND MACHINE LEARNING. Lecture 6: Data preprocessing and model selection Lecturer: Simone Scardapane

DATA MINING AND MACHINE LEARNING. Lecture 6: Data preprocessing and model selection Lecturer: Simone Scardapane DATA MINING AND MACHINE LEARNING Lecture 6: Data preprocessing and model selection Lecturer: Simone Scardapane Academic Year 2016/2017 Table of contents Data preprocessing Feature normalization Missing

More information

Introduction to Mobile Ad hoc Networks (MANETs)

Introduction to Mobile Ad hoc Networks (MANETs) Introduction to Mobile Ad hoc Networks (MANETs) 1 Overview of Ad hoc Network Communication between various devices makes it possible to provide unique and innovative services. Although this inter-device

More information

An Introduction to Dynamic Simulation Modeling

An Introduction to Dynamic Simulation Modeling Esri International User Conference San Diego, CA Technical Workshops ****************** An Introduction to Dynamic Simulation Modeling Kevin M. Johnston Shitij Mehta Outline Model types - Descriptive versus

More information

Using Hidden Semi-Markov Models for Effective Online Failure Prediction

Using Hidden Semi-Markov Models for Effective Online Failure Prediction Using Hidden Semi-Markov Models for Effective Online Failure Prediction Felix Salfner and Miroslaw Malek SRDS 2007 Presented by: Fahad Arshad School of Electrical and Computer Engineering Purdue University

More information

Anomaly Detection in Cyber Physical Systems

Anomaly Detection in Cyber Physical Systems Anomaly Detection in Cyber Physical Systems Maggie Cheng Illinois Institute of Technology December 11, 2018 IEEE Big Data Conference Seattle, WA Outline Introduction Outlier Detection Sequential Change

More information

Hierarchy of knowledge BIG DATA 9/7/2017. Architecture

Hierarchy of knowledge BIG DATA 9/7/2017. Architecture BIG DATA Architecture Hierarchy of knowledge Data: Element (fact, figure, etc.) which is basic information that can be to be based on decisions, reasoning, research and which is treated by the human or

More information

Cisco Crosswork Network Automation

Cisco Crosswork Network Automation Cisco Crosswork Network Introduction Communication Service Providers (CSPs) are at an inflexion point. Digitization and virtualization continue to disrupt the way services are configured and delivered.

More information

Swarming in NuPIC. September 6, 2013

Swarming in NuPIC. September 6, 2013 Swarming in NuPIC September 6, 2013 What swarming does Automatically determines the best CLA model for a given dataset Determines which components should go into the model (encoders, spatial & temporal

More information

Section 4 General Factorial Tutorials

Section 4 General Factorial Tutorials Section 4 General Factorial Tutorials General Factorial Part One: Categorical Introduction Design-Ease software version 6 offers a General Factorial option on the Factorial tab. If you completed the One

More information

Numenta Node Algorithms Guide NuPIC 1.7

Numenta Node Algorithms Guide NuPIC 1.7 1 NuPIC 1.7 includes early implementations of the second generation of the Numenta HTM learning algorithms. These algorithms are available as two node types: SpatialPoolerNode and TemporalPoolerNode. This

More information

INTERNATIONAL TELECOMMUNICATION UNION STUDY GROUP 12 DELAYED CONTRIBUTION 95. Behaviour of data modems in lossy, packet-based transport systems

INTERNATIONAL TELECOMMUNICATION UNION STUDY GROUP 12 DELAYED CONTRIBUTION 95. Behaviour of data modems in lossy, packet-based transport systems INTERNATIONAL TELECOMMUNICATION UNION TELECOMMUNICATION STANDARDIZATION SECTOR STUDY PERIOD 2005-2008 English only Original: English Questions: 11, 13/12 Geneva, 17-21 October 2005 Source: Title: STUDY

More information

CHAPTER 5 OUTLIER CLEANING AND SENSOR DATA AGGREGATION USING MODIFIED Z-SCORE TECHNIQUE

CHAPTER 5 OUTLIER CLEANING AND SENSOR DATA AGGREGATION USING MODIFIED Z-SCORE TECHNIQUE 72 CHAPTER 5 OUTLIER CLEANING AND SENSOR DATA AGGREGATION USING MODIFIED Z-SCORE TECHNIQUE 5.1 INTRODUCTION The outlier detection is an important preprocessing routine that is required to ensure robustness

More information

Healthcare IT A Monitoring Primer

Healthcare IT A Monitoring Primer Healthcare IT A Monitoring Primer Published: February 2019 PAGE 1 OF 13 Contents Introduction... 3 The Healthcare IT Environment.... 4 Traditional IT... 4 Healthcare Systems.... 4 Healthcare Data Format

More information

Strategic Briefing Paper Big Data

Strategic Briefing Paper Big Data Strategic Briefing Paper Big Data The promise of Big Data is improved competitiveness, reduced cost and minimized risk by taking better decisions. This requires affordable solution architectures which

More information

Artificial Neuron Modelling Based on Wave Shape

Artificial Neuron Modelling Based on Wave Shape Artificial Neuron Modelling Based on Wave Shape Kieran Greer, Distributed Computing Systems, Belfast, UK. http://distributedcomputingsystems.co.uk Version 1.2 Abstract This paper describes a new model

More information

HW #4. Hierarchical Temporal Memory

HW #4. Hierarchical Temporal Memory HW #4. Hierarchical Temporal Memory Implement episodic memory using an HTM network. Skeleton codes are available from http://rit.kaist.ac.kr/home/ee788-2017 Using the code, you are required to store and

More information

ZKLWHýSDSHU. 3UHð)DLOXUHý:DUUDQW\ý 0LQLPL]LQJý8QSODQQHGý'RZQWLPH. +3ý 1HW6HUYHUý 0DQDJHPHQW. Executive Summary. A Closer Look

ZKLWHýSDSHU. 3UHð)DLOXUHý:DUUDQW\ý 0LQLPL]LQJý8QSODQQHGý'RZQWLPH. +3ý 1HW6HUYHUý 0DQDJHPHQW. Executive Summary. A Closer Look 3UHð)DLOXUHý:DUUDQW\ý 0LQLPL]LQJý8QSODQQHGý'RZQWLPH ZKLWHýSDSHU Executive Summary The Hewlett-Packard Pre-Failure Warranty 1 helps you run your business with less downtime. It extends the advantage of

More information

March 10 11, 2015 San Jose

March 10 11, 2015 San Jose March 10 11, 2015 San Jose Health monitoring & predictive analytics To lower the TCO in a datacenter Christian B. Madsen & Andrei Khurshudov Engineering Manager & Sr. Director Seagate Technology christian.b.madsen@seagate.com

More information

Intelligent Routing Platform

Intelligent Routing Platform WHITE PAPER Bring Intelligence to your Network Copyright 2018 Noction Inc. Table of Contents 1. Executive Summary...3 2. The Challenge of a Multi-Homed Environment...4 3. Network Congestion and Blackouts...4

More information

International Journal of Data Mining & Knowledge Management Process (IJDKP) Vol.7, No.3, May Dr.Zakea Il-Agure and Mr.Hicham Noureddine Itani

International Journal of Data Mining & Knowledge Management Process (IJDKP) Vol.7, No.3, May Dr.Zakea Il-Agure and Mr.Hicham Noureddine Itani LINK MINING PROCESS Dr.Zakea Il-Agure and Mr.Hicham Noureddine Itani Higher Colleges of Technology, United Arab Emirates ABSTRACT Many data mining and knowledge discovery methodologies and process models

More information

Remotely Sensed Image Processing Service Automatic Composition

Remotely Sensed Image Processing Service Automatic Composition Remotely Sensed Image Processing Service Automatic Composition Xiaoxia Yang Supervised by Qing Zhu State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University

More information

FP7 NEMESYS Project: Advances on Mobile Network Security

FP7 NEMESYS Project: Advances on Mobile Network Security Enhanced Network Security for Seamless Service Provisioning in the Smart Mobile Ecosystem FP7 NEMESYS Project: Advances on Mobile Network Security Elina Theodoropoulou R&D Projects Section Manager etheodorop@cosmote.gr

More information

Application of Characteristic Function Method in Target Detection

Application of Characteristic Function Method in Target Detection Application of Characteristic Function Method in Target Detection Mohammad H Marhaban and Josef Kittler Centre for Vision, Speech and Signal Processing University of Surrey Surrey, GU2 7XH, UK eep5mm@ee.surrey.ac.uk

More information

PARALLEL AND DISTRIBUTED PLATFORM FOR PLUG-AND-PLAY AGENT-BASED SIMULATIONS. Wentong CAI

PARALLEL AND DISTRIBUTED PLATFORM FOR PLUG-AND-PLAY AGENT-BASED SIMULATIONS. Wentong CAI PARALLEL AND DISTRIBUTED PLATFORM FOR PLUG-AND-PLAY AGENT-BASED SIMULATIONS Wentong CAI Parallel & Distributed Computing Centre School of Computer Engineering Nanyang Technological University Singapore

More information

Detection of Missing Values from Big Data of Self Adaptive Energy Systems

Detection of Missing Values from Big Data of Self Adaptive Energy Systems Detection of Missing Values from Big Data of Self Adaptive Energy Systems MVD tool detect missing values in timeseries energy data Muhammad Nabeel Computer Science Department, SST University of Management

More information

High Availability and Redundant Operation

High Availability and Redundant Operation This chapter describes the high availability and redundancy features of the Cisco ASR 9000 Series Routers. Features Overview, page 1 High Availability Router Operations, page 1 Power Supply Redundancy,

More information

CMSC 341 Lecture 16/17 Hashing, Parts 1 & 2

CMSC 341 Lecture 16/17 Hashing, Parts 1 & 2 CMSC 341 Lecture 16/17 Hashing, Parts 1 & 2 Prof. John Park Based on slides from previous iterations of this course Today s Topics Overview Uses and motivations of hash tables Major concerns with hash

More information

Big Data Methods. Chapter 5: Machine learning. Big Data Methods, Chapter 5, Slide 1

Big Data Methods. Chapter 5: Machine learning. Big Data Methods, Chapter 5, Slide 1 Big Data Methods Chapter 5: Machine learning Big Data Methods, Chapter 5, Slide 1 5.1 Introduction to machine learning What is machine learning? Concerned with the study and development of algorithms that

More information

Tree-based methods for classification and regression

Tree-based methods for classification and regression Tree-based methods for classification and regression Ryan Tibshirani Data Mining: 36-462/36-662 April 11 2013 Optional reading: ISL 8.1, ESL 9.2 1 Tree-based methods Tree-based based methods for predicting

More information

BITS F464: MACHINE LEARNING

BITS F464: MACHINE LEARNING BITS F464: MACHINE LEARNING Lecture-16: Decision Tree (contd.) + Random Forest Dr. Kamlesh Tiwari Assistant Professor Department of Computer Science and Information Systems Engineering, BITS Pilani, Rajasthan-333031

More information

Best Practice for Creation and Maintenance of a SAS Infrastructure

Best Practice for Creation and Maintenance of a SAS Infrastructure Paper 2501-2015 Best Practice for Creation and Maintenance of a SAS Infrastructure Paul Thomas, ASUP Ltd. ABSTRACT The advantage of using metadata to control and maintain data and access to data on databases,

More information

Introduction to Hadoop and MapReduce

Introduction to Hadoop and MapReduce Introduction to Hadoop and MapReduce Antonino Virgillito THE CONTRACTOR IS ACTING UNDER A FRAMEWORK CONTRACT CONCLUDED WITH THE COMMISSION Large-scale Computation Traditional solutions for computing large

More information

IBM Data Science Experience White paper. SparkR. Transforming R into a tool for big data analytics

IBM Data Science Experience White paper. SparkR. Transforming R into a tool for big data analytics IBM Data Science Experience White paper R Transforming R into a tool for big data analytics 2 R Executive summary This white paper introduces R, a package for the R statistical programming language that

More information

Paging algorithms. CS 241 February 10, Copyright : University of Illinois CS 241 Staff 1

Paging algorithms. CS 241 February 10, Copyright : University of Illinois CS 241 Staff 1 Paging algorithms CS 241 February 10, 2012 Copyright : University of Illinois CS 241 Staff 1 Announcements MP2 due Tuesday Fabulous Prizes Wednesday! 2 Paging On heavily-loaded systems, memory can fill

More information

Review on ichat: Inter Cache Hardware Assistant Data Transfer for Heterogeneous Chip Multiprocessors. By: Anvesh Polepalli Raj Muchhala

Review on ichat: Inter Cache Hardware Assistant Data Transfer for Heterogeneous Chip Multiprocessors. By: Anvesh Polepalli Raj Muchhala Review on ichat: Inter Cache Hardware Assistant Data Transfer for Heterogeneous Chip Multiprocessors By: Anvesh Polepalli Raj Muchhala Introduction Integrating CPU and GPU into a single chip for performance

More information

Automate Transform Analyze

Automate Transform Analyze Competitive Intelligence 2.0 Turning the Web s Big Data into Big Insights Automate Transform Analyze Introduction Today, the web continues to grow at a dizzying pace. There are more than 1 billion websites

More information

Predicting and Stopping Account Takeover Fraud. A PatternEx Solution Guide

Predicting and Stopping Account Takeover Fraud. A PatternEx Solution Guide Predicting and Stopping Account Takeover Fraud A PatternEx Solution Guide Problem Every major breach fundamentally comes down to criminals obtaining legitimate credentials. Using valid ID s and passwords,

More information

Data Intensive Scalable Computing

Data Intensive Scalable Computing Data Intensive Scalable Computing Randal E. Bryant Carnegie Mellon University http://www.cs.cmu.edu/~bryant Examples of Big Data Sources Wal-Mart 267 million items/day, sold at 6,000 stores HP built them

More information

FIREWALL CLEANUP RECOMMENDATIONS

FIREWALL CLEANUP RECOMMENDATIONS WHITEPAPER FIREWALL CLEANUP RECOMMENDATIONS CONSIDERATIONS FOR IMPROVED FIREWALL EFFICIENCY, BETTER SECURITY, AND REDUCED POLICY COMPLEXITY EXECUTIVE SUMMARY Firewalls are designed to provide access control.

More information

LOW LATENCY DATA DISTRIBUTION IN CAPITAL MARKETS: GETTING IT RIGHT

LOW LATENCY DATA DISTRIBUTION IN CAPITAL MARKETS: GETTING IT RIGHT LOW LATENCY DATA DISTRIBUTION IN CAPITAL MARKETS: GETTING IT RIGHT PATRICK KUSTER Head of Business Development, Enterprise Capabilities, Thomson Reuters +358 (40) 840 7788; patrick.kuster@thomsonreuters.com

More information

Predicting Service Outage Using Machine Learning Techniques. HPE Innovation Center

Predicting Service Outage Using Machine Learning Techniques. HPE Innovation Center Predicting Service Outage Using Machine Learning Techniques HPE Innovation Center HPE Innovation Center - Our AI Expertise Sense Learn Comprehend Act Computer Vision Machine Learning Natural Language Processing

More information

Software Security IV: Fuzzing

Software Security IV: Fuzzing 1 Software Security IV: Fuzzing Chengyu Song Slides modified from Dawn Song 2 Administrivia Homework1 Due: Friday Oct 27 11:59pm Questions regarding reading materials Talk Security R&D in a Security Company:

More information

The Power of the Crowd

The Power of the Crowd WHITE PAPER The Power of the Crowd SUMMARY With the shift to Software-as-a-Service and Cloud nearly complete, organizations can optimize their end user experience and network operations with the power

More information

EECS 122: Introduction to Computer Networks Resource Management and QoS. Quality of Service (QoS)

EECS 122: Introduction to Computer Networks Resource Management and QoS. Quality of Service (QoS) EECS 122: Introduction to Computer Networks Resource Management and QoS Computer Science Division Department of Electrical Engineering and Computer Sciences University of California, Berkeley Berkeley,

More information

Acquisition Description Exploration Examination Understanding what data is collected. Characterizing properties of data.

Acquisition Description Exploration Examination Understanding what data is collected. Characterizing properties of data. Summary Statistics Acquisition Description Exploration Examination what data is collected Characterizing properties of data. Exploring the data distribution(s). Identifying data quality problems. Selecting

More information

Test Oracles. Test Oracle

Test Oracles. Test Oracle Encontro Brasileiro de Testes de Software April 23, 2010 Douglas Hoffman, BACS, MBA, MSEE, ASQ-CSQE, ASQ-CMQ/OE, ASQ Fellow Software Quality Methods, LLC. (SQM) www.softwarequalitymethods.com doug.hoffman@acm.org

More information

Echo Cancellation in VoIP Applications. Jeffrey B. Holton. Texas Instruments June, 2002

Echo Cancellation in VoIP Applications. Jeffrey B. Holton. Texas Instruments June, 2002 Echo Cancellation in VoIP Applications Jeffrey B. Holton Texas Instruments June, 2002 Abstract Voice-over-IP networks present a significant cost savings over traditional PSTN solutions. However, the technologies

More information

Machine Learning for Pre-emptive Identification of Performance Problems in UNIX Servers Helen Cunningham

Machine Learning for Pre-emptive Identification of Performance Problems in UNIX Servers Helen Cunningham Final Report for cs229: Machine Learning for Pre-emptive Identification of Performance Problems in UNIX Servers Helen Cunningham Abstract. The goal of this work is to use machine learning to understand

More information

ACELLENT SOFTWARE CATALOG

ACELLENT SOFTWARE CATALOG ACELLENT SOFTWARE CATALOG Software Acellent's software works in tandem with our sensors and hardware to detect and characterize structural anomalies in metals and composites due to the presence of cracks,

More information

Measuring VDI Fitness and User Experience Technical White Paper

Measuring VDI Fitness and User Experience Technical White Paper Measuring VDI Fitness and User Experience Technical White Paper 3600 Mansell Road Suite 200 Alpharetta, GA 30022 866.914.9665 main 678.397.0339 fax info@liquidwarelabs.com www.liquidwarelabs.com Table

More information

CHAPTER 6 STATISTICAL MODELING OF REAL WORLD CLOUD ENVIRONMENT FOR RELIABILITY AND ITS EFFECT ON ENERGY AND PERFORMANCE

CHAPTER 6 STATISTICAL MODELING OF REAL WORLD CLOUD ENVIRONMENT FOR RELIABILITY AND ITS EFFECT ON ENERGY AND PERFORMANCE 143 CHAPTER 6 STATISTICAL MODELING OF REAL WORLD CLOUD ENVIRONMENT FOR RELIABILITY AND ITS EFFECT ON ENERGY AND PERFORMANCE 6.1 INTRODUCTION This chapter mainly focuses on how to handle the inherent unreliability

More information

High Availability Using Fault Tolerance in the SAN. Wendy Betts, IBM Mark Fleming, IBM

High Availability Using Fault Tolerance in the SAN. Wendy Betts, IBM Mark Fleming, IBM High Availability Using Fault Tolerance in the SAN Wendy Betts, IBM Mark Fleming, IBM SNIA Legal Notice The material contained in this tutorial is copyrighted by the SNIA. Member companies and individual

More information

Data Driven Networks

Data Driven Networks Data Driven Networks Is it possible for to Learn the control planes of networks and applications? Operators specify what they want, and the system learns how to deliver CAN WE LEARN THE CONTROL PLANE OF

More information

Three requirements for reducing performance issues and unplanned downtime in any data center

Three requirements for reducing performance issues and unplanned downtime in any data center Three requirements for reducing performance issues and unplanned downtime in any data center DARRYL FUJITA TECHNICAL SOFTWARE SOLUTIONS SPECIALIST HITACHI DATA SYSTEMS How Big Is The Cost Of Unplanned

More information

CS Review. Prof. Clarkson Spring 2017

CS Review. Prof. Clarkson Spring 2017 CS 5430 Review Prof. Clarkson Spring 2017 Recall: Audit logs Recording: what to log what not to log how to log locally remotely how to protect the log Reviewing: manual exploration automated analysis MANUAL

More information

Kaggle See Click Fix Model Description

Kaggle See Click Fix Model Description Kaggle See Click Fix Model Description BY: Miroslaw Horbal & Bryan Gregory LOCATION: Waterloo, Ont, Canada & Dallas, TX CONTACT : miroslaw@gmail.com & bryan.gregory1@gmail.com CONTEST: See Click Predict

More information

Using the Forest to See the Trees: Context-based Object Recognition

Using the Forest to See the Trees: Context-based Object Recognition Using the Forest to See the Trees: Context-based Object Recognition Bill Freeman Joint work with Antonio Torralba and Kevin Murphy Computer Science and Artificial Intelligence Laboratory MIT A computer

More information

Anomaly Detection Fault Tolerance Anticipation

Anomaly Detection Fault Tolerance Anticipation Anomaly Detection Fault Tolerance Anticipation Patterns John Allspaw SVP, Tech Ops Qcon London 2012 Four Cornerstones Erik Hollnagel (Anticipation) (Response) Knowing Knowing Knowing Knowing What What

More information

Structured Light II. Thanks to Ronen Gvili, Szymon Rusinkiewicz and Maks Ovsjanikov

Structured Light II. Thanks to Ronen Gvili, Szymon Rusinkiewicz and Maks Ovsjanikov Structured Light II Johannes Köhler Johannes.koehler@dfki.de Thanks to Ronen Gvili, Szymon Rusinkiewicz and Maks Ovsjanikov Introduction Previous lecture: Structured Light I Active Scanning Camera/emitter

More information

ENGINEERING SENSOR NETWORKS FOR LONG- TERM ENVIRONMENTAL MONITORING. Guest lecturer: Doug Carlson

ENGINEERING SENSOR NETWORKS FOR LONG- TERM ENVIRONMENTAL MONITORING. Guest lecturer: Doug Carlson ENGINEERING SENSOR NETWORKS FOR LONG- TERM ENVIRONMENTAL MONITORING Guest lecturer: Doug Carlson 2 Hi. Who are you? I m Doug Carlson, one of Dr. Terzis s PhD students. I have been doing research on wireless

More information

The trend to virtualisation

The trend to virtualisation The trend to virtualisation Key concepts Desktop virtualisation describes the process of separating or abstracting a personal desktop environment (including its operating system, applications, user profile

More information

White Paper: VANTIQ Digital Twin Architecture

White Paper: VANTIQ Digital Twin Architecture Vantiq White Paper www.vantiq.com White Paper: VANTIQ Digital Twin Architecture By Paul Butterworth November 2017 TABLE OF CONTENTS Introduction... 3 Digital Twins... 3 Definition... 3 Examples... 5 Logical

More information

Large Scale Data Analysis for Policy

Large Scale Data Analysis for Policy Large Scale Data Analysis for Policy 90-866, Fall 2012 Lecture 9: Anomaly and Outlier Detection Parts of this lecture were adapted from Banerjee et al., Anomaly Detection: A Tutorial, presented at SDM

More information

FUZZY INFERENCE SYSTEMS

FUZZY INFERENCE SYSTEMS CHAPTER-IV FUZZY INFERENCE SYSTEMS Fuzzy inference is the process of formulating the mapping from a given input to an output using fuzzy logic. The mapping then provides a basis from which decisions can

More information

Data Preprocessing. Slides by: Shree Jaswal

Data Preprocessing. Slides by: Shree Jaswal Data Preprocessing Slides by: Shree Jaswal Topics to be covered Why Preprocessing? Data Cleaning; Data Integration; Data Reduction: Attribute subset selection, Histograms, Clustering and Sampling; Data

More information

Sampling Large Graphs for Anticipatory Analysis

Sampling Large Graphs for Anticipatory Analysis Sampling Large Graphs for Anticipatory Analysis Lauren Edwards*, Luke Johnson, Maja Milosavljevic, Vijay Gadepally, Benjamin A. Miller IEEE High Performance Extreme Computing Conference September 16, 2015

More information

Some economical principles

Some economical principles Hints on capacity planning (and other approaches) Andrea Bianco Telecommunication Network Group firstname.lastname@polito.it http://www.telematica.polito.it/ Some economical principles Assume users have

More information

Page Replacement. (and other virtual memory policies) Kevin Webb Swarthmore College March 27, 2018

Page Replacement. (and other virtual memory policies) Kevin Webb Swarthmore College March 27, 2018 Page Replacement (and other virtual memory policies) Kevin Webb Swarthmore College March 27, 2018 Today s Goals Making virtual memory virtual : incorporating disk backing. Explore page replacement policies

More information

Automated Detection of Individual Anomalies and General Behavioral Changes in Time Series Metrics

Automated Detection of Individual Anomalies and General Behavioral Changes in Time Series Metrics Automated Detection of Individual Anomalies and General Behavioral Changes in Time Series Metrics Abstract Tifany Yung, Renato Martins Groupon, Inc. tiyung@groupon.com, rmartins@groupon.com Metrics provide

More information

Hardware Acceleration of Inference Computing The Numenta HTM Algorithm

Hardware Acceleration of Inference Computing The Numenta HTM Algorithm Hardware Acceleration of Inference Computing The Numenta HTM Algorithm Dan Hammerstrom Electrical And Computer Engineering 5/13/11 1 My research program focuses on specialized computer architectures for

More information

A CAN-Based Architecture for Highly Reliable Communication Systems

A CAN-Based Architecture for Highly Reliable Communication Systems A CAN-Based Architecture for Highly Reliable Communication Systems H. Hilmer Prof. Dr.-Ing. H.-D. Kochs Gerhard-Mercator-Universität Duisburg, Germany E. Dittmar ABB Network Control and Protection, Ladenburg,

More information

Augmented Spatial Pooling

Augmented Spatial Pooling Augmented Spatial Pooling John Thornton, Andrew Srbic, Linda Main, and Mahsa Chitsaz Institute for Integrated and Intelligent Systems, Griffith University, QLD, Australia j.thornton@griffith.edu.au {andrew.srbic,linda.main,mahsa.chitsaz}@griffithuni.edu.au

More information

Medical practice: diagnostics, treatment and surgery in supercomputing centers

Medical practice: diagnostics, treatment and surgery in supercomputing centers International Advanced Research Workshop on High Performance Computing from Clouds and Big Data to Exascale and Beyond Medical practice: diagnostics, treatment and surgery in supercomputing centers Prof.

More information

CS229 Lecture notes. Raphael John Lamarre Townshend

CS229 Lecture notes. Raphael John Lamarre Townshend CS229 Lecture notes Raphael John Lamarre Townshend Decision Trees We now turn our attention to decision trees, a simple yet flexible class of algorithms. We will first consider the non-linear, region-based

More information

Slides for Data Mining by I. H. Witten and E. Frank

Slides for Data Mining by I. H. Witten and E. Frank Slides for Data Mining by I. H. Witten and E. Frank 7 Engineering the input and output Attribute selection Scheme-independent, scheme-specific Attribute discretization Unsupervised, supervised, error-

More information

Big Data Insights Using Analytics

Big Data Insights Using Analytics Governors State University OPUS Open Portal to University Scholarship All Capstone Projects Student Capstone Projects Fall 2015 Big Data Insights Using Analytics Naga Krishna Reddy Muppidi Governors State

More information

Cognitive augmented routing system and its standardisation path

Cognitive augmented routing system and its standardisation path Cognitive augmented routing system and its standardisation path ETSI Future Network Technologies Workshop Dimitri Papadimitriou, Bernard Sales Alcatel-Lucent, Bell Labs March, 2010 Self-Adaptive (top-down)

More information

Programming with MPI

Programming with MPI Programming with MPI p. 1/?? Programming with MPI Miscellaneous Guidelines Nick Maclaren Computing Service nmm1@cam.ac.uk, ext. 34761 March 2010 Programming with MPI p. 2/?? Summary This is a miscellaneous

More information

Smart Manufacturing Enterprise

Smart Manufacturing Enterprise Smart Manufacturing Enterprise The Plant of the Future Marco Gamba, Smart Manufacturing Project Manager Fabio Beda, Technical Sales Consultant Smart Manufacturing An evolution which raises from a need

More information

Weka ( )

Weka (  ) Weka ( http://www.cs.waikato.ac.nz/ml/weka/ ) The phases in which classifier s design can be divided are reflected in WEKA s Explorer structure: Data pre-processing (filtering) and representation Supervised

More information

BUSINESS CONTINUITY: THE PROFIT SCENARIO

BUSINESS CONTINUITY: THE PROFIT SCENARIO WHITE PAPER BUSINESS CONTINUITY: THE PROFIT SCENARIO THE BENEFITS OF A COMPREHENSIVE BUSINESS CONTINUITY STRATEGY FOR INCREASED OPPORTUNITY Organizational data is the DNA of a business it makes your operation

More information

NEC Virtualized Evolved Packet Core vepc

NEC Virtualized Evolved Packet Core vepc TE-524262 NEC Virtualized Evolved Packet Core vepc Design Concepts and Benefits INDEX Leading the transformation into Mobile Packet Core Virtualization P.3 vepc System Architecture Overview P.4 Elastic

More information

Datacenter replication solution with quasardb

Datacenter replication solution with quasardb Datacenter replication solution with quasardb Technical positioning paper April 2017 Release v1.3 www.quasardb.net Contact: sales@quasardb.net Quasardb A datacenter survival guide quasardb INTRODUCTION

More information

Transforming Transport Infrastructure with GPU- Accelerated Machine Learning Yang Lu and Shaun Howell

Transforming Transport Infrastructure with GPU- Accelerated Machine Learning Yang Lu and Shaun Howell Transforming Transport Infrastructure with GPU- Accelerated Machine Learning Yang Lu and Shaun Howell 11 th Oct 2018 2 Contents Our Vision Of Smarter Transport Company introduction and journey so far Advanced

More information

Lab Testing Summary Report

Lab Testing Summary Report Key findings and conclusions: Lab Testing Summary Report July 2008 Report 080711 Vendor Tested: Product Tested: WinEyeQ Product Category: VoIP Analysis and Management Tools WinEyeQ was rated Best in Test

More information

Course: Advanced Software Engineering. academic year: Lecture 14: Software Dependability

Course: Advanced Software Engineering. academic year: Lecture 14: Software Dependability Course: Advanced Software Engineering academic year: 2011-2012 Lecture 14: Software Dependability Lecturer: Vittorio Cortellessa Computer Science Department University of L'Aquila - Italy vittorio.cortellessa@di.univaq.it

More information

From Correlation to Causation: Active Delay Injection for Service Dependency Detection

From Correlation to Causation: Active Delay Injection for Service Dependency Detection From Correlation to Causation: Active Delay Injection for Service Dependency Detection Christopher Kruegel Computer Security Group ARO MURI Meeting ICSI, Berkeley, November 15, 2012 Correlation Engine

More information

Capacity planning and.

Capacity planning and. Hints on capacity planning (and other approaches) Andrea Bianco Telecommunication Network Group firstname.lastname@polito.it http://www.telematica.polito.it/ Some economical principles Assume users have

More information

Adaptive Background Mixture Models for Real-Time Tracking

Adaptive Background Mixture Models for Real-Time Tracking Adaptive Background Mixture Models for Real-Time Tracking Chris Stauffer and W.E.L Grimson CVPR 1998 Brendan Morris http://www.ee.unlv.edu/~b1morris/ecg782/ 2 Motivation Video monitoring and surveillance

More information

Performance Assurance in Virtualized Data Centers

Performance Assurance in Virtualized Data Centers Autonomic Provisioning with Self-Adaptive Neural Fuzzy Control for End-to-end Delay Guarantee Palden Lama Xiaobo Zhou Department of Computer Science University of Colorado at Colorado Springs Performance

More information

COSC160: Detection and Classification. Jeremy Bolton, PhD Assistant Teaching Professor

COSC160: Detection and Classification. Jeremy Bolton, PhD Assistant Teaching Professor COSC160: Detection and Classification Jeremy Bolton, PhD Assistant Teaching Professor Outline I. Problem I. Strategies II. Features for training III. Using spatial information? IV. Reducing dimensionality

More information

Intelligent Embedded Systems

Intelligent Embedded Systems Advanced Research Intelligent Embedded Systems Cognitive Fault Detection and Diagnosis in Sensor-based Embedded Systems Intelligence for Embedded Systems Ph. D. and Master Course Manuel Roveri Politecnico

More information

Improved Detection of Low-Profile Probes and Denial-of-Service Attacks*

Improved Detection of Low-Profile Probes and Denial-of-Service Attacks* Improved Detection of Low-Profile Probes and Denial-of-Service Attacks* William W. Streilein Rob K. Cunningham, Seth E. Webster Workshop on Statistical and Machine Learning Techniques in Computer Intrusion

More information