Analytics: The Machine Learning Advantage

The Machine Learning Advantage

Machine learning is, to keep it simple, an algorithm developed to note changes in data and evolve in it’s design to accommodate the new findings. As applied to predictive analytics, this feature has wide ranging impact on the activities normally undertaken to develop, test, and refine an algorithm for a given purpose.

Sophisticated pattern recognition – Along with noting relationships, the Yottamine Predictive Platform can determine the type and quantify as well. This is not just happening with key, or even secondary variables, but on every relationship that takes part in the pattern. This feature delineates irrelevant data as well, which provides the benefits of mitigating pre-processing requirements and accelerating processing. Since the solution has an ordering or ranking capability, the key variables self-identify as a part of the processing.

Intelligent decisions – Along with the capability to note irrelevant data, and rank the relative importance of variables, Yottamine will make decisions either aided by the user or not. This becomes apparent when modeling to predict a rare event. The solution can distinguish subclasses and make determinations on what data should be included and which shouldn’t with very little instruction.

Self modifying – Clearly, being able to tweak, add, or drop different aspects of an algorithm to better typify the data is a time saver. However, as this is also taking place to accommodate minor variables and sub-classes, so time demands are being held in check while the accuracy of the algorithm, and its ability to predict are significantly improved.

Multiple iterations – As the model becomes more refined, YPP tests multiple iterations to produce a final version that delivers the highest level of accuracy while maintaining the best fit to the data.

Machine Learning Theory Analytics + Cloud

Yottamine is the first company to leverage the cloud as a source of high performance computing to drive advanced predictive analytics solutions. This enabled Yottamine to fully apply the most advanced thinking in Machine Learning Theory without compromise or cutting corners. The benefits of this approach make Yottamine the most intelligent choice for nearly all corporate Big Data needs.

Power – Having all the power you need, when you need it, means you have the power to drive the most intelligent predictive analytics solution available. Power also means running the volumes of data required for precise answers to your questions. No matter how you view Big Data, the power is there to extract the knowledge you need from your biggest and most challenging data.

Flexibility – Modeling and data mining projects come in all shapes and sizes, but typically, in setting up an in-house solution, cost considerations limit processing power to projects of average size. All of a sudden, planning isn’t based on average and projects can become something far more than average.

Computing without Infrastructure – No Infrastructure limitations or a complicated distributed computing architecture to design, implement, and constantly monitor. Instead, have high performance computing power available when you need it, for any number of projects necessary.

Cost Control – Meeting your BIG DATA requirements on a small data budget. Or let other business units manage their requirements on their small budget!

Kernel Based Learning and Support Vector Machines

Support Vector Machines (SVMs) represent the latest advancement in machine learning theory and deliver state of the art performance in numerous high value applications. Full scale SVMs have been difficult to put into production because such powerful procedures are resource intensive to compute.

The core of SVMs consist of a maximal margin hyperplane (also referred to as the decision boundary) and the use of a kernel function to transform the original data set into a high dimensional space (a space for all possible combinations of predictive variables). The margin maximization principle has not only been proven mathematically to deliver robust and predictable performance on unseen data, but also been shown to deliver state of the art performance on many real-world application.


A kernel function transforms the data set from original input space into a high dimensional feature space in which, highly nonlinear relationships between the factors or attributes are qualified and examined using the margin maximization principle. This allows SVMs to solve and uncovered highly complex relationships, and deliver state of the art performance.

The Yottamine scientific team consist of globally recognized experts in machine learning theory and applications concentrating in SVMs. The high performance SVMs developed by Yottamine are parallelized and fully leverage the cloud as an elastic compute resource.