Data Visualization With GGobi -- A Hands-On Tutorial*
GGobi is an award-winning open source data analysis and visualization tool. This hands-on tutorial will focus on visual methods for classification, also known as supervised machine learning. Install packages and sample data tested on Ubuntu, openSUSE and Fedora will be provided.
There are five steps in the data analysis process:
1. The Problem Statement
2. Data Preparation
3. Exploratory Data Analysis
4. Quantitative Analysis
GGobi is an award-winning open source data analysis and visualization tool with advanced capabilities that can be applied to all of these steps. GGobi can help remove outliers, deal with missing data points, develop both supervised and unsupervised machine learning algorithms, determine whether the “patterns” one sees are really there, and even help analyze social networks.
This tutorial will focus on using visual methods in GGobi for classification, also known as supervised learning. Participants are encouraged to bring a laptop computer running one of the major community Linux distributions to experience GGobi first hand. Install packages and sample data tested on Ubuntu, openSUSE and Fedora will be provided.
M. Edward (Ed) Borasky is an applied mathematician and computer scientist with interests ranging from computer performance engineering to algorithmic composition and synthesis of music. His open source interests include Linux, Ruby, PostgreSQL and R.
- Title: Linux Server Profiling
- Track: BoF
- Room: Broadway
- Time: 7:00 – 8:30pm
A number of open source tools exist that make profiling Linux servers easier. These tools include traditional Unix utilities like “sar” and “iostat”, but they also include some tools that go deep into the processors and I/O subsystems.
- Speakers: Ed Borasky