Introduction to machine
learning in Elastic
Search. Observe. Protect.
Tom Grabowski, ML Product Manager
Camilla Montonen, Elastic ML data scientist
Jenny Morris, Solutions Architect
June 25, 2020
Meet Our Speakers
Housekeeping & Logistics
● Attendees are automatically muted when joining Zoom webinar
● Q+A will be at the end of the webinar
● Ask questions for us in the Zoom chat during the webinar
○ Chat settings To: All panelists and attendees
○ Ask more questions on our discuss forum: discuss.elastic.co
● Recording will be available after the webinar and on the event page
Elastic is a search company.
Scale RelevanceSpeed
Three Solutions.
Elastic Enterprise Search Elastic SecurityElastic Observability
One Stack.
Elastic Enterprise Search Elastic SecurityElastic Observability
Kibana
Elasticsearch
Beats Logstash
Deploy Anywhere.
Elastic Enterprise Search Elastic SecurityElastic Observability
Kibana
Elasticsearch
Beats Logstash
Elastic
Cloud
Elastic Cloud
on Kubernetes
Elastic Cloud
Enterprise
Self-Managed
Elastic Stack
Elastic Machine Learning
Operationalize and Simplify
data science
Time Series Anomaly Detection Data Frame Analysis
Machine Learning Anomaly Detection
Unsupervised machine learning
Automatically detect anomalies, outliers from
group, and rare events
Sophisticated ML Job UI
Interactive views of model and anomaly scoring
Root cause analysis
Report on factors influencing anomalies
On-Demand Forecasting
Forecast out time series metrics
10 years development & Industry leading technology
111
Predict
Expected value @ 15:05 = 1859
Learn Operationalize
Solution integrations
Anomaly detection within the Solutions workflow
APM
Logs
SIEM
Uptime
Anomaly Detection
Demo
Machine Learning expanding use cases
Unsupervised and supervised machine learning
Unsupervised
Supervised
Data Driven:
Pattern
Recognition
Labelled data
for
Learning and
Predicting
Anomaly Detection
Outlier Detection
Forecasting
Language ID
Fraud Detection
User classification
15
What behaviour can you learn from to make predictions?
• ML in search
– What language is this document written in?
– How can I boost search relevance for named entities?
– What search results are most relevant based on click-through rate?
• Observability
– What users or hosts are outliers?
– How can I Classify alerts and route them to the right team?
– What customers are likely to churn?
• Security
– How can I identify malicious domain names generated by DGAs?
– How can I classify activity as originating from a device type (e.g. router version)?
Why build your own ML models?
Using Supervised Learning
for DGA Identification
Elastic is a Search Company.
www.elastic.co
Thank You
SLED Virtual User Group
Learn more about Elastic and represent your community at the next
Elastic Virtual Group - July 23 @ 2 PM ET
Monitoring and preventing threats as employees transition
from home to office
As employees return to offices or continue to work from home, the attack surface
increases exponentially. How will you protect your network and infrastructures
with new users, changes to roles and permissions, and high volumes of attack
attempts? Get insights from Salt Lake County for protecting their systems with
insights on more comprehensive monitoring and logging, improved alerting, and
more!
View Details
Questions?
Elastic is a search company.
Tom Grabowski
ML Product Manager
Elastic
Camilla Montonen
ML Data Scientist
Elastic
Jenny Morris
Solutions Architect
Elastic

Introduction to machine learning using Elastic

  • 1.
    Introduction to machine learningin Elastic Search. Observe. Protect. Tom Grabowski, ML Product Manager Camilla Montonen, Elastic ML data scientist Jenny Morris, Solutions Architect June 25, 2020
  • 2.
  • 3.
    Housekeeping & Logistics ●Attendees are automatically muted when joining Zoom webinar ● Q+A will be at the end of the webinar ● Ask questions for us in the Zoom chat during the webinar ○ Chat settings To: All panelists and attendees ○ Ask more questions on our discuss forum: discuss.elastic.co ● Recording will be available after the webinar and on the event page
  • 4.
    Elastic is asearch company. Scale RelevanceSpeed
  • 5.
    Three Solutions. Elastic EnterpriseSearch Elastic SecurityElastic Observability
  • 6.
    One Stack. Elastic EnterpriseSearch Elastic SecurityElastic Observability Kibana Elasticsearch Beats Logstash
  • 7.
    Deploy Anywhere. Elastic EnterpriseSearch Elastic SecurityElastic Observability Kibana Elasticsearch Beats Logstash Elastic Cloud Elastic Cloud on Kubernetes Elastic Cloud Enterprise Self-Managed Elastic Stack
  • 8.
    Elastic Machine Learning Operationalizeand Simplify data science
  • 9.
    Time Series AnomalyDetection Data Frame Analysis
  • 10.
    Machine Learning AnomalyDetection Unsupervised machine learning Automatically detect anomalies, outliers from group, and rare events Sophisticated ML Job UI Interactive views of model and anomaly scoring Root cause analysis Report on factors influencing anomalies On-Demand Forecasting Forecast out time series metrics 10 years development & Industry leading technology
  • 11.
    111 Predict Expected value @15:05 = 1859 Learn Operationalize
  • 12.
    Solution integrations Anomaly detectionwithin the Solutions workflow APM Logs SIEM Uptime
  • 13.
  • 14.
    Machine Learning expandinguse cases Unsupervised and supervised machine learning Unsupervised Supervised Data Driven: Pattern Recognition Labelled data for Learning and Predicting Anomaly Detection Outlier Detection Forecasting Language ID Fraud Detection User classification
  • 15.
    15 What behaviour canyou learn from to make predictions? • ML in search – What language is this document written in? – How can I boost search relevance for named entities? – What search results are most relevant based on click-through rate? • Observability – What users or hosts are outliers? – How can I Classify alerts and route them to the right team? – What customers are likely to churn? • Security – How can I identify malicious domain names generated by DGAs? – How can I classify activity as originating from a device type (e.g. router version)? Why build your own ML models?
  • 16.
    Using Supervised Learning forDGA Identification
  • 17.
    Elastic is aSearch Company. www.elastic.co Thank You
  • 18.
    SLED Virtual UserGroup Learn more about Elastic and represent your community at the next Elastic Virtual Group - July 23 @ 2 PM ET Monitoring and preventing threats as employees transition from home to office As employees return to offices or continue to work from home, the attack surface increases exponentially. How will you protect your network and infrastructures with new users, changes to roles and permissions, and high volumes of attack attempts? Get insights from Salt Lake County for protecting their systems with insights on more comprehensive monitoring and logging, improved alerting, and more! View Details
  • 19.
  • 20.
    Elastic is asearch company.
  • 21.
    Tom Grabowski ML ProductManager Elastic Camilla Montonen ML Data Scientist Elastic Jenny Morris Solutions Architect Elastic