Blogs

Understand MPP in-memory code execution with Regression at Kognitio webinar

With many more professionals looking to improve the efficiency of their big data analytics operations using advanced in-memory computing technologies, understanding how to make the most of these solutions will be crucial to the success of a scheme.

This is why Kognitio is currently hosting a series of in-depth online workshops that are aiming to inform users of best practices for taking advantage of the powerful tools. And the next such session will be taking place on March 19th, with a focus on how Kognitio's solutions can help businesses perform regression analytics in order to gain more useful insight from their data.

Beginning at 8am PST/11am EDT/3pm GMT, the session will last for two hours, which should give enough time for attendees to gain thorough answers to any questions they might have about the use of regression algorithms in in-memory massively parallel processing (MPP) environments. Previous workshops that took this approach proved to be highly instructive, receiving great accolades from all those who attended, so the upcoming session will seek to offer more of this highly-advanced level of insight.

It will focus on using R to conduct these analytics activities, but as we recognise some professionals may feel more comfortable with other options, the techniques we detail will also be applicable to other stats packages and programming languages. The workshop will be led by Dr Sharon Kirkham, head data scientist at Kognitio and director of the Kognitio Analytics Center of Excellence

Among the areas that are set to be covered during the workshop, we will review how the Kognitio scripting interface works, with examination of a simple regression algorithm as an example. Then it will also illustrate the 'divide and conquer' approach that has been proven to work so well on parallel platforms. 

However, helping users to understand how to cut out complex coding and use the languages they are familiar with to manage their data flows and produce results is only the start. The workshop will also detail how this can be integrated into your wider organization.

Attendees will hear about how to organize their information into a single script, which can then be used to gain different levels of insight, depending on the unique requirements of a business' users.

Finally, the event will discuss some of the concepts and thought processes that professionals will need to consider when looking at full parallelization of tasks such as regression. This will be demonstrated to attendees with an example in linear regression.

As the workshop is designed to be an open session, questions will be welcome throughout the event, so if a user has a specific query of how they might apply regression algorithms to their own MPP solutions, this presents a perfect chance to ask and hear the insight of Kognitio's experts.

More information about the topics up for discussion is available on the Kognitio User Forum for External Scripting. Here, developers, programmers and data scientists will be able to find modules available to help educate themselves on the technology.

Please click here to register for this virtual event.