Data Analysis and Visualization

Data Analysis includes computational methods to extract, process, clean, transform, and visualize large, complex data sets. Additionally, it includes statistical anaysis using both parametric and non-parametric techniques. JDSAT personnel have proven experience in supporting service headquarters of the US Navy and Marine Corps by employing a broad range of data & statistical techniques.

Data Mining includes a rich set of machine learning techniques to fit highly complex data models for accurate prediction. Common techniques include: Classification & Regression Trees, Neural Networks, Logistic & Linear Regression, Non-Linear Regression, Data Clustering & Visualization, Feature/Variable Reduction, Feature/Variable Importance Ranking, Pattern Matching, and Anomaly Detection.

 

Common Applications

  • Feature Importance – Which product features are most important for predicting sales?
  • Credit Scoring – How likely is one to pay on time?
  • Fraud Detection – Where are the anomalies in the data?
  • Classification – Which features predict sucess in an occupation?
  • Segment Markets – What price/incentives appeal to different people?
  • Data Clustering & Visualization
  • Statistical Sampling, Stratification

 

Relevant Experience

  • Enterprise military/medical manpower analysis
  • Survey Design, Analysis, and Program Management
  • Classification efficiency of military occupations
  • Detection of potential ASVAB test “cheaters”
  • Testing of gender-based differences in promotions

datamining