GraphLab, a machine learning and graph computation startup, announced the general availability of their flagship product GraphLab Create 1.0 at the Strata + HadoopWorld 2014 conference on Wednesday.
The software platform brings large-scale machine learning capabilities including predictive analytics to data scientists working in organisations.
Carlos Guestrin, CEO of GraphLab, explains, “Being creative with data given the limitations and complexities of current tools is a data scientist’s greatest obstacle to productivity.”
“The availability of GraphLab Create will empower data scientists, software developers and engineers alike to leverage all of their data stores for building the data products that can make significant impacts to their business,” he said.
The machine learning platform is tailor made for developers and data scientists with functional programming skills and some basic understanding of data science, reports their press release, enabling ‘prototyping and scaling of ideas from inspiration to production’, making deployment and integration easy by turning machine learning models into predictive services consumable by front-end web applications.
Example services include recommenders, fraud detectors or customer churn predictors.
The announcement of public availability also included the unveiling of a raft of new features. These include:
- Predictive Services – scalable, fault-tolerant, and high performing predictive service deployments enabling integration with front-end applications
- Deep Learning – combined with GraphLab Create™ image analysis tools, the Deep Learning package enables accurate and in-depth understanding of images and videos, making quick work of importing and preprocessing millions of images as well as numeric data
- Boosted Trees – support for this popular class of algorithms for robust and accurate regression and classification tasks can easily scale up to large datasets that do not fit into memory
- Visualization – users can visualize the status and health of offline jobs deployed in various environments including local, Hadoop Clusters and EC2.
Adobe, Zillow and Crosswise are a few of the companies that use GraphLab Create and Data science training program Zipfian Academy has also incorporated the software into its curriculum so data scientists can take advantage of its speed, ease of use, and ability to handle large data sets.
GraphLab was founded in 2009 as an open source project by PhD., Carlos Guestrin. Now available in three full-featured editions Discover, Develop and Deploy GraphLab is also working with a number of customers for Enterprise-wide use of GraphLab Create.
Read more here.
(Image credit: GraphLab)
The Author
