Yottamine Predictive Service (YPS) brings unprecedented prediction speed, scale, accuracy and price-performance to organizations of any size. Based on industry accepted web services, it includes comprehensive capabilities for importing and applying models in a real-world setting. It is designed to allow users to take full advantage of scalable on-demand cloud computing, and eliminate the high costs of a dedicated infrastructure. The Yottamine Predictive Service allows for building models or making predictions in two simple steps. Via integration with scalable cloud computing it provides high speed and efficiency. It also conforms with the SSL industry security standard and can exports to PMML those models that are supported by the standard.
Getting started with Yottamine Predictive Services is very straightforward for the data scientist, as they can connect and control Yottamine Predictive Web Services using R programming language via YottamineR package. Data scientists will be able to get up and running very quickly, and simply using a R session on their laptop to harness the power of Yottamine Predictive Service (YPS) on public or private cloud-based cluster for building models from big data and perform prediction in real-time..
To watch a short video demonstration on how Yottamine Predictive Service and YottamineR work, click here.
Yottamine Predictive Service consists of the four following web services which gives data scientist maximum flexibility for solving machine learning problem in most applications.
1. MODEL BUILDING SERVICE: Build models based on the imported files from Storage Service or existing files from R session, S3 in hadoop hdfs format.
- Model Building Simplified: Build a model on the cloud from a single web service call; we automatically calculate the number of instances and EC2 power needed.
- Real-time interaction: Check your model building progress or stop building a model at any time.
- You own your models: Convert your models to industry standard PMML format for use with other applications. (NOTE: Only for models supported by PMML Standard)
- Maximum flexibility: Build as many large or small models as you like without worrying about scalability or additional cost of adding more hardware.
- Highly Accurate Models: Build highly accurate models by utilizing the power of cloud computing and get high-performance algorithms within a short period of time.
- Add value to your existing services: Perform real-time scoring by line or query from your MySQL database.
- Freedom: Adjustable EC2 computer power based on instance type.
- Unparalleled flexibility: Run as many predictions on as many models as you like; more instances will be started in order to fit the next model into memory.
3. STORAGE SERVICE: Import data for model building and running predictions
- Access your existing data from anywhere: Import your data outside the S3 domain or data inside your Amazon S3 native format to our hosted S3 or your S3 in hadoop hdfs format.
- Easy export: Export your models outside the S3 domain or data to the Amazon S3 native format in your S3.
- Manage your data: Perform S3 file operations such as list, copy, rename, move, and delete in both Amazon S3 native or hadoop hdfs format.
4. AUTHENTICATION SERVICE: Required for logging into the web service and retrieve settings from client management console
- Token based: Token based authentication to reduce passing sensitive password information for each web service call.
- Smart timeout: System auto logs web service session out based on NTP time servers when there are no more web service calls from the client
For latest price and package information, please contact us.