gm-spacagna/sceval
ScEval, pronounced skev-al, is a highly Scalable evaluation framework using Spark for Scala
Discovered public repositories for gm-spacagna in the GitHub catalog.
ScEval, pronounced skev-al, is a highly Scalable evaluation framework using Spark for Scala
Cascading and Scalding wrapper for HBase with advanced read features
Mirror of Apache Spark
ID2209 Homework1
TopTal project: ToDo Task List Manager
TunUp: A Parallel Genetic Auto-Tuning system for Big Data Machine Learning algorithms
ID2210-Lab-Group6
The year 2010, the information society relies more and more on software agents. Mobile agents gradually take over time consuming information search tasks. Supplied with a particular task and strategy to find information, mobile agents travel through the Internet to visit information sources. They act to some extend autonomously. The Mobile Information Agent Inc. (MIA) provides infrastructure for information retrieving. This infrastructure contains information source stations and light-weight portable agent computers. Each agent computer contains equipment needed for Internet based communication. It also has a Java processor, Java OS as well as Java runtime environment. You are hired by the MIA to develop the application software for information source stations as well as mobile agents. The MIA requires that the software agents must be mobile, social, and intelligent. You must fulfill the minimum requirements for the software, but you are also told that innovations and additions of your own will be rewarded.