User Tools

Site Tools


16.08 - Machine Learning Ensemble in MapReduce for Predictive Analytics

Project - Summary


The main goal of this project is to design algorithms which can fully leverage the MapReduce framework and machine learning techniques to make reliable predictions. The objectives include:

  • Developing an ensemble based machine learning approach for predictive analytics
  • Parallelizing algorithms of action rules mining and contrast set mining for MapReduce
  • Utilizing multiple models to produce more robust predictions

Project - Team

Team Member Role Email Phone Number Academic Site/IAB
Jian Chen PI Not available Not available UL Lafayette
Jennifer Lavergne PI Not available Not available UL Lafayette
Ryan Benton Co-PI Not available Not available UL Lafayette
Cheng-Yuan Ke Graduate Student Not available Not available UL Lafayette
Cirish Chalamalasetti Graduate Student Not available Not available UL Lafayette

Project - Impact and Uses/Benefits

Predictive analytics enables information extraction from big data (both real-time and historical data) to forecast what might happen in the future with an acceptable level of reliability, including what-if scenarios and risk assessment. Machine learning methods are widely used in many domains for prediction but not many of them are MapReduce enabled. We developed a machine learning ensemble predictive analytics system on MapReduce platform. Our system offers many benefits for the IAB members. These benefits include:

  1. Combines Anomaly Detection with Action Association Rules with transitions that describe interesting changes in the data over time
  2. Utulizes Contract Sets Mining to discover hotspots and emerging/declining trends
  3. Predicts data points a year in advance for a specified topic or area

In addition, the system can also be modified minimally to explore multiple different areas. Examples include disease outbreaks prediction (public health), forecasting on emerging markets, cross-selling, advertisement optimization (marketing), fighting terrorism, crimes, cyber security (security), and identifying high-risk fraud candidates (fraud detection).

Project - Deep Dive

Project - Documents

FilenameFilesizeLast modified
16.08_year_5_poster.pptx764.8 KiB2019/08/22 13:21
16.08_year_5_presentation.pdf678.1 KiB2019/08/22 13:21
16.08_year_5_final_report.pdf2.3 MiB2019/08/22 10:47
projects/year5/16.08.txt · Last modified: 2019/08/22 10:49 by sally.johnson