Experiment with Splunk Machine Learning

[melissap]

View our Tech Talk on Experiment with Splunk Machine Learning to learn  how you can use your data with the Splunk Machine Learning Toolkit app available on splunkbase for outlier and anomaly detection, predictive analytics and data clustering.

• Learn about how to model your Splunk data using machine learning
• Engage in a live demonstration of the Splunk Machine Learning Toolkit
• Get access to the latest resources on Machine Learning in Splunk

Check out our  Machine Learning Toolkit conversations in Splunk Answers community for more!

[melissap]

Hey everyone! We had some great questions during this Tech Talk in May. 

Recapping for all!

Q: What ML toolkit is being used on the demo?

A: MLTK 5.2

Q: How could I use machine learning for sets of data that start at different times each day, for example if I wanted to monitor my heart rate on a bike ride every day, but I start my ride at different times each day. I want to predict my future fitness levels based on my improved heart rate

A: If this activity is once a day, then do a span=1 day , and move ahead with forecasting

Q: How to know what best model and algorithm to be used on my use case since there are so many available?

A: Smart Assistant comes with the best algorithm suited for the use case. if you're not a data scientist, you don't need to choose from the available options..just use the new smart assistant

Q: Do you provide special service that can help us guide our own use cases?

A: Check this out : https://www.splunk.com/en_us/software/splunk-enterprise/machine-learning.html#MLTK-FAQ

Q: Do you have use case using ITSI and Machine Learning together?

A: ITSI has predictive analytics which uses MLTK.

Q: Will Splunk come up with a machine learning certificate in the future?

A: There is a data analytics course using Machine learning in Splunk today and it comes with a certification

Q: Is it compatible with Splunk Enterprise version 7.3.3?

A: No..you have to be on Splunk 8.0 and above but forecasting and outlier detection is available in MLTK 4.5 which is compatible with 7.3.3

Q: How to find out the Threshold frequency for Outlier detection.

A: Its an output from algorithm..Look at density function algorithm docs in Splunk MLTK

Q: What is the maximum amount of data (csv, kv, etc.) that the MLTK has been tested with?

A: Please check the MLTK performance app for algorithm reports

[melissap]

Recapping for all!

Q: What ML toolkit is being used on the demo?

A: MLTK 5.2

Q: How could I use machine learning for sets of data that start at different times each day, for example if I wanted to monitor my heart rate on a bike ride every day, but I start my ride at different times each day. I want to predict my future fitness levels based on my improved heart rate

A: If this activity is once a day, then do a span=1 day , and move ahead with forecasting

Q: How to know what best model and algorithm to be used on my use case since there are so many available?

A: Smart Assistant comes with the best algorithm suited for the use case. if you're not a data scientist, you don't need to choose from the available options..just use the new smart assistant

Q: Do you provide special service that can help us guide our own use cases?

A: Check this out : https://www.splunk.com/en_us/software/splunk-enterprise/machine-learning.html#MLTK-FAQ

Q: Do you have use case using ITSI and Machine Learning together?

A: ITSI has predictive analytics which uses MLTK.

Q: Will Splunk come up with a machine learning certificate in the future?

A: There is a data analytics course using Machine learning in Splunk today and it comes with a certification

Q: Is it compatible with Splunk Enterprise version 7.3.3?

A: No..you have to be on Splunk 8.0 and above but forecasting and outlier detection is available in MLTK 4.5 which is compatible with 7.3.3

Q: How to find out the Threshold frequency for Outlier detection.

A: Its an output from algorithm..Look at density function algorithm docs in Splunk MLTK

Q: What is the maximum amount of data (csv, kv, etc.) that the MLTK has been tested with?

A: Please check the MLTK performance app for algorithm reports

[melissap]

Here are some additional resources for your Machine Learning journey:

• MLTK App on Splunkbase
• Documentation
• Machine Learning Blogs
• conf papers
  
  • The Machine Learning Toolkit in Action FN 1470
  • The New Experiment Experience in the Splunk Machine Learning Toolkit FN 1553
  • Predict Real World Outage using Splunk MLTK IT 1722
  • The Two Most Common Machine Learning Solutions Everyone Needs to Know FN 1213