Machine Learning, Human Learning and Big Data

Woelk Photo By Darrell Woelk

Some new things are brewing at Elastic Knowledge and we will start to fill you in through a series of blogs by the Elastic Knowledge staff.

We started Elastic Knowledge in 2002 with a focus on the integration of knowledge management and online learning. While we will continue our commitment to human learning, there are some exciting things happening in the world of machine learning and big data. These developments have encouraged us to expand our focus also into the areas of machine learning and its relationship with knowledge management.

Machine learning techniques for discovering new patterns in data and making predictions using those patterns have been around for a while. Effectively applying these techniques and associated algorithms requires a deep understanding of statistics and is both an art and a science. New capabilities to amass vast amounts of data have vastly improved machine learning capabilities, simply due to the fact that the computer has so many more examples to learn from. Thus, big data is enabling machine learning in more practical areas, as organizations now are generating petabytes of data (data exhaust) from their day to day operations. This data contains hidden knowledge about their customers, products and processes that may hold the key to making the organization more efficient or profitable.

Applying machine learning to analyze big data is not easy. The available big data tools are open source, poorly documented and immature. Machine learning algorithms that perform well on small data sets may fail miserably on larger data sets. Additionally, knowing which business problems and questions might be addressed by a particular machine learning technique on a given set of data is itself both an art and a science. The individuals that have the skill set to do these tasks are called Data Scientists.

Elastic Knowledge consultants now include Data Scientists for hire. These consultants have the expertise to help companies who are just starting to amass large amounts of data, or have data that they do not know how to analyze, to understand how they can make sense of what they have. We have the both the theoretical statistical knowledge and the practical data management experience to help an organization analyze its business problem, investigate the data needed to solve the problem, decide what machine learning techniques to apply, apply the techniques and analyze and explain the results.

Through an upcoming series of blogs, we will explain our strategies and methodologies from three points of view: Business, Science and Technology. Our goal is to help explain the business problems that can be addressed using a learning approach, the science behind the learning approach and the technologies to implement the learning approach. We look forward to sharing our knowledge and experience with you and look forward to your feedback to help us work together to keep up with this emerging and fast-moving field.