Apr 19, 2018 7:56:00 AM

Splunk Machine Learning vs. Google Machine Learning

Deploying machine learning in your environment can be a scary proposition. However, as the technology becomes more available, the barrier of entry has been lowered for companies looking to leverage this powerful tool. Two of August Schell’s partners, Splunk and Google, have developed platforms that give users the power to automate Machine Learning in their enterprise with simple, easy-to-use interfaces.

Read on for a brief overview of each developer’s offerings, along with their comparative advantages.

Splunk

The Machine Learning Toolkit (MLTK)

One of Splunk MLTK’s primary advantages over the competition is its swift integration with your already-existing Splunk deployment. Simply download their free app off of Splunkbase and you’re ready to go.

The toolkit features a number of pre-built models, allowing you to operate right out of the box. The Showcase menu is packed with examples to help you understand the function of each model and its best practices. Users can do things like detecting outliers in Bitcoin transactions, or predicting internet traffic based on the time of day.

User Behavior Analytics (UBA)

Splunk’s other core machine learning offering is its User Behavior Analytics. It monitors your network data to establish “normal” parameters and then triggers alerts and anomalies whenever security issues or threats arise. For example, say a user on your network renames a number of protected files that they don’t normally access. UBA detects this activity and tags it for the analyst’s review

Insight Engines

August Schell is also a proud partner of Insight Engines, a third-party Splunk application that allows users to make Splunk queries in human-readable terms. Insight Engines allows organizations to distribute the workload of Splunk across more of their team, so even those without knowledge of the Splunk Processing Language can get value out of the system.

Cost

Splunk’s pricing is based on a licensing model. You pay upfront for a certain amount of data indexed per month, along with their Support Service. Prices start at $150/Gb/month, and decrease as you hit higher tiers. This means your business needs to plan for your data usage before making a decision.

google-machine-learning

Google

Tensorflow

Almost all of Google’s machine learning products work with their in-house, open-source Python library, Tensorflow. All of the models you make are of your own design, using their language. This offers greater flexibility than Splunk, but at the cost of losing some of the out-of-the-box functionality.

Google Cloud Platform

Google’s machine learning engine is centered within the rest of their Cloud Platform, so you gain access to the tight integration that Google products are known for. All of the Cloud Platform is easily accessible from the same single pane of glass, allowing you to look at your entire pipeline from one dashboard.

Multimedia Capability

The Cloud Platform is much more equipped than Splunk to handle a variety of media. It contains Image Recognition and Video Processing software that automatically parse and annotate user content.

Cost

Google works with a pay-as-you-use system, charging users based on how much data they’re consuming. There are different charges based on the service used, as well. For the Machine Learning Engine in particular, prices start around $0.27/hr for the smallest jobs, and scales from there. Additional charges are added for each node used in the job.

Ultimately, the benefit to either of these systems comes down to the needs of your environment. Splunk has the relative advantage in networked environments, where you already have a monitored feed and want to add some additional machine learning functionality to find out more about your data. It’s simple, easy to use, and easy to install. Google thrives in situations involving a larger range of media, such as images, video, etc. The Cloud Platform is also a much farther reaching tool, performing virtualization and maintaining backend for your business in addition to machine learning.

Need some help figuring out which Machine Learning system is right for your business? Feel free to contact August Schell online or by phone at (301)-838-9470.