Over a million developers have joined DZone.
{{announcement.body}}
{{announcement.title}}

Exploring the Intersection of Machine Learning and Analytics

DZone's Guide to

Exploring the Intersection of Machine Learning and Analytics

This article presents a summary of how machine learning algorithms can impact business analytics. Learn about Anomaly Detection, Natural Language Processing, and much more!

· Big Data Zone
Free Resource

Free O'Reilly eBook: Learn how to architect always-on apps that scale. Brought to you by Mesosphere DC/OS–the premier platform for containers and big data.

When I was a young boy I saw the classic movie “2001 A Space Odyssey“ with HAL, the voice interactive computer system that bordered on AI, and that sparked in me, a lifelong interest and career in IT.

Today we are seeing devices that are starting to provide the beginnings of that same functionality like the Amazon Echo, Dot, or Google Home. It’s one thing to sit in your living room and call out to the air “Alexa, how old is Matt Damon” or “Alexa, play 'The Logical Song' by Supertramp”, it’s another when your 6 year old is having a conversation with Alexa and orders a bunch of things from Amazon and it is quite another when you are trying to find ways to use it in the office to make your company more productive.

Machine Learning

The difference between Alexa and HAL is pretty dramatic, but at the core of them both, and AI in general, is Machine Learning. As Guy Levy-Yurista, Sisense Head of Product, described in this recent blog post:  “Sisense employs machine learning as a core element of its In-Chip™ data processing algorithms….We call it query recycling – breaking queries into smaller blocks that are later reassembled to answer future queries: if user A asks a completely new question such as ‘what was our average deal size last year?’ and user B later asks ‘what is our year-over-year growth in sales?’, the second user will receive their answer faster – because the system has already performed some of aggregations needed to answer their question.

This isn’t OLAP, it is a learning algorithm that grows smarter and more efficient over time and as more unique queries accumulate. It learns to identify the reusable chunks within each query, and to use these as a knowledge base for future reference”.

Let’s think a bit about how machine learning algorithms can impact business analytics. In general, once you’ve established what your pressure points are, those are the same metrics you will often return to. That means that those queries, and related queries will get very fast. Typical examples would be something like revenue figures for some period of time. What it also does is provide a base to grow beyond a simple dashboard. How about some of these ideas.

Anomaly Detection

There are obvious and not so obvious ways to make use of a feature like an anomaly detection. If sales fluctuate within 5% in a day, maybe that is normal, but if you are over 15%, that is cause for concern. You can look at your dashboards and find this information, but with anomaly detection, you can trigger events through the Sisense ecosystem. Maybe this is sending you a text or an email or using Zapier to integrate with a 3rd party application, it could even be a message in Slack.

Consider some other ideas of where you could use these kinds of alerts. How about law enforcement? Maybe there is a seasonality to certain types of crimes or within a certain geographic region. If you have a sudden spike in drug related activity for example, wouldn’t an immediate notice be useful? Then you get out your dashboards and start drilling into what is happening and maybe tie certain other types of information from the municipality, maybe a new apartment complex just opened or a school shut down. Getting that big picture really helps you look for patterns. With SiSense Pulse, you are able to create all this smart alerting with anomaly detection.

Natural Language Interaction

Natural Language Interaction

All this machine learning gives us other great ways to interact with our data as well. With the SiSense chatbot, you can carry on a chat conversation from a messaging system like Slack or Skype, in which you engage in a natural language conversation with Sisense to get insights. You might type something like “Summarize sales by region for the fourth quarter of 2016” and Sisense will present you with a written response that might also be accompanied by a graph to further elaborate on the data provided.

From this point, you have information to ask other questions, such as drilling down on a region to look at the store or salesperson performance. You could even do this while you’re on the run with the Sisense mobile app.

We started off talking about HAL and Alexa, which are voice response system. What if you could say to your Amazon Echo “Alexa, what was the gross sales figure for yesterday?” and get a response? People might wonder who you are talking to when they walk by your office, but this ability is now a reality with SiSense. You can issue voice commands to an Echo and retrieve information in real time, to your questions. As you become more accustomed to this method of interacting with business analytics, the more uses you are going to find.

What started as a seemingly basic implementation of machine learning for performance purposes, has led to the core of an entirely new way to perform analytics. You’re really only limited by your imagination.

Easily deploy & scale your data pipelines in clicks. Run Spark, Kafka, Cassandra + more on shared infrastructure and blow away your data silos. Learn how with Mesosphere DC/OS.

Topics:
big data ,anomaly detection ,natural language processing

Published at DZone with permission of Shawn Gordon, DZone MVB. See the original article here.

Opinions expressed by DZone contributors are their own.

THE DZONE NEWSLETTER

Dev Resources & Solutions Straight to Your Inbox

Thanks for subscribing!

Awesome! Check your inbox to verify your email so you can start receiving the latest in tech news and resources.

X

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}