Oxdata last week released the second generation of its H20 prediction engine for users of Hadoop, R and Excel. H20 uses in-memory on Hadoop and provides parallel and distributed algorithms on big data, which it claims runs 100 times faster than other predictive analytics providers.
Oxdata CEO and co-founder SriSatish Ambati said H2O Fluid Vector is simple and easy-to-use and deploy from R, Excel and the Hadoop ecosystem. “We bring the power of Google-scale machine learning and modeling without sampling to the rest of the world."
The firm describes itself as an open-source machine-learning and predictive analytics company for big data.
Madan Sheina, Ovum’s lead analyst for information management, said the release highlights the clever alternative approaches to solve complex data analytics issues in large data scaling environments at speed – without throwing more expensive hardware at the problem.
Sheina said one of those approaches is machine-learning – which sits at the edge of statistics and computer science – that focuses on the development of fast and efficient algorithms for real-time processing of data and to deliver accurate predictions. “Machine learning techniques are complex and differ from more traditional statistical techniques.”
He said the company’s statement that the release is “the industry’s fastest and simplest prediction engine for Hadoop data” is a bold claim, but it’s something they’ve been working at for a long time. He sees three design traits that stand out: speed of analysis, scale of data and simplicity of use.
Oxdata said the statistical analysis engine is tightly integrated with R, which uses the Hadoop Distributed File System (HDFS) as its storage platform. Users can ingest data from Microsoft Excel, the RStudio integrated developer environment, SQL, NoSQL, S3 or HDFS using a REST API.
Analysys Mason research director Patrick Kelly said the upgrade was a great technology enabler for Hadoop, but noted “the real value for businesses is applying big data in real time to achieve specific business outcomes. This requires business rule logic and speed. H2O is the speed enabler but still requires business logic algorithms.”