YottamineOP for Hadoop

Slider 02 Profit Comparison 72dpiHigher Business Return

YottamineOP leverages big data stored in your Hadoop Cluster and the power of Yottamine’s unique patent-pending Optimized Prediction machine learning technology to bring businesses higher ROI. Unlike most of the machine learning products, YottamineOP specifically designed for applications such as churn analysis, fraud detection and risk modelling, where the occurrence of desirable (such as purchasing) or undesirable (e.g. commit a fraud)  outcome of a business event is only a small portion of all the possible events. Results from real-world business applications have proven YottamineOP to be superior than typical machine learning techniques such as Gradient Boosting Machine and Support Vector Machines on improving business return. To learn more please read our whitepaper, “Higher Business Return with Optimized Prediction” .

Designed for Speed and Runs on Big DataData Hub 02 (1)

Unlike most advanced machine learning algorithms that are slow to run. YottamineOP is developed from the ground up to utilize today’s big data platform Hadoop and Spark. For ultimate performance, YottamineOP can also offload the most compute intensive part of the model building to native code based MPI framework, bringing supercomputer performance to your existing Hadoop Cluster for building Optimized Prediction model. To learn more about the speed and scale this can achieve, please check out our Scalability Benchmark done by a Hadoop Cluster in AWS cloud.

Reduced Time from Data to Prediction

YottamineOP is based on advanced machine learning theory that can take all available features into consideration without sacrificing model accuracy. It reduces the amount of iteration required for the data scientist in figuring out which variables are important. In other words, it takes full advantage of the depth and volume of big data stored in your Enterprise Data Hub.

Automated Parameter EstimationData Hub 03

No more complex parameter tuning is required, YottamineOP will automatically estimate hard to tune parameters for machine learning algorithm to maximize profit.

RlogoFamiliar R Programming Interface

YottamineOP has a simple to use R interface that allows you to easily build Optimized Prediction models without having to learn a new language.


Certified to run on Cloudera’s Enterprise Data Hub

logo_Cloudera_CertifiedYottamineOP is certified by Cloudera to run on CDH 5.x and work with its advanced security feature. It is easy to install on your existing secure CDH Hadoop Cluster. Please check out online documentation to find out more.

To start a free evaluation, please contact sales@yottamine.com.

Want to integrate Optimized Prediction into your existing apps? Please contact oemsales@yottamine.com.

Want to see if YottamineOP can bring higher ROI to your existing Machine Learning project, but does not have the proper hardware? Check out our cloud based product that does not required any installation of software here.