Home Contact Us Site Map

Predictive Analytics

Predictive Analytics encompasses a variety of techniques from statistics and Data Mining that analyze current and historical data to make predictions about future events. Our computer scientists have significant hands-on experience in conceiving, developing and deploying diagnostic and prognostic solutions for Threat Assessment and Mitigation, Drug Discovery, Integrated Vehicle Health Management (IVHM), Cyber-Security, Wind Turbine monitoring and other Anomaly/Fault Detection applications.

At MAC, we emphasize the three major components of Predictive Analytics:

  • All aspects of Data Mining / Pattern Recognition / Machine Learning including detection, classification, regression, clustering and anomaly detection using a variety of powerful paradigms
  • Data Warehousing for handling massive historical datasets and modern data streams, such as network traffic, that can generate vast quantities of data in short periods of time
  • Interactive Visualization techniques that enable drill-down and facilitate hypothesis formation

Data Mining

The Data Exploitation Group's scientists have extensive experience in designing and implementing Pattern Recognition / Machine Learning / Data Mining applications. We choose the right paradigm for the problem at hand based on our expertise. The following table lists some of the techniques we have used in the past to solve challenging problems in Industry:
Learn More

Neural Networks

Decision Trees and Ensembles

Support Vector Machines

K Nearest Neighbors

Clustering Techniques

Bayesian Learning

Evolutionary Algorithms

Hidden Markov Models

Self Organizing Maps (SOMs)

Data Warehousing

An extremely important aspect of Predictive Analytics is the collection and maintenance of large data repositories to enable historical analysis as well as to handle modern monitoring applications capable of generating terabytes of data in short periods. MAC researchers have experience in designing schema and developing proper strategies for efficiently storing records using Database platforms such as MySQL, Postgres and Oracle.


Choosing the proper techniques for interactively summarizing, filtering and visualizing massive datasets can make the difference between understanding it and being confused by it. At MAC, we have considerable understanding of the depth and breadth of visualization techniques and the know-how about what algorithm to use when. Following is an abbreviated list of techniques that we often use to shed light on datasets:
  • Frequency Plots
  • Scatter Plots
  • Treemaps
  • Tree Pies
  • Bubble Plots
  • Bubble Pies
  • Bar Charts
  • Polar Bar Charts
  • Structured Network Plots
  • Map Plots
  • 3D Plots and Projection
  • Pixel-based Views
MAC has the tools and talent to deliver robust Predictive Analytics solutions.

Copyright © 2000-2014 Michigan Aerospace Corporation. All rights reserved.
Send any questions or comments to