Data Consolidation of Disparate Procurement Data Sources for Correlated Performance-Based Acquisition Decision Support
MetadataShow full item record
Frank Kendall, then Under Secretary of Defense for Acquisition, Technology and Logistics, released the first defense acquisition system performance report in June 2013. This report focused primarily on performance related to the collective outcomes of Major Defense Acquisition Programs (MDAPs), but additionally explored various descriptive dimensions and acquisition approaches of the same (Kendall, 2013). Each annual report builds on the work previously conducted, and focuses on data-driven analysis relying on statistical techniques to identify trends that improve the defense acquisition communityﾒs insights into how contract incentives are motivating better contractor/vendor performance (Kendal, 2016). Nevertheless, large amounts of data (in modern jargon, ﾓBig Dataﾔ) are now available for research in the area of defense acquisition. Over the past several years, changes in electronic commerce have increased the amounts of both structured and unstructured data availableﾗboth in runtime and archived environments. This electronic data, from a variety of different acquisition agencies, can be obtained by a variety of means and used for a multitude of purposes (Snider et al., 2014). Traditional statistical and trend analysis methods thus far have been primarily relied upon to explore trends and test metrics in the sets of acquisition data at hand. Sometimes, spreadsheets of linear regression correlation are employed, or, in some more modern applications, multivariate structural equation models via scientific applications such as SPSS and AMOS are leveraged for their ability to evaluate complex variable relationships, such as nested or recursive if-then patterns (Byrne, 2016). However, not only are todayﾒs modern datasets large in magnitude, they are also large in variety and complexity (Gartner, 2013). Furthermore, to address this state of data, new statistical modeling techniques, more powerful than before, have had to be created. This is due to the older methods finding difficulty with some of the size problems Big Data represents, such as privacy and security concerns (Parms, 2017). Thankfully, computer power necessary to employ the modern techniques is less expensive today, the software near free, and the storage capacities available now yield bewildering capacities at a fingertip, and with amazingly fast access speed. In fact, these performance parameters appear to continue along a Mooreﾒs trend line against critical opposition (Magee, Basnet, Funk, & Benson, 2015). Presently, one of the more interesting of the new statistical modeling techniques is neural network algorithm machine learning. Neural network modeling involves utilizing a ﾓpowerful computational data model that is able to capture and represent input/output relationships.ﾔ This model was developed out of the desire to create artificial intelligence systems capable of completing functions that were previously executed solely by the human brain. One benefit of using neural network modeling lies with its capacity to display and comprehend both linear and non-linear relationships from the data to which it is supplied (NeuroSolutions, 2015). Research Question Because ﾓBig Dataﾔ is present in the Defense Acquisition Business space, and, because the demand to critically understand real cause-and-effect relationships between variables within that data is persistent from the Acquisition community, this paperﾒs research question is, Can a neural network modeling technique be confidently relied upon to meaningfully explore variable relationships within acquisition business datasets? Because, if it is, then any question may be reasonably asked by anyone of such a dataset; and, via the neural network-enabled tool, the answers they receive will come with scientific statistical confidence as to whether they can be trusted as interesting or useful answers.1 In order to explore this research question, the study opted to use business data on contractor performance and attempted to isolate predictive variables from past performance information predictive of good performance.
RightsThis publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.
NPS Report NumberSYM-AM-17-044
Showing items related by title, author, creator and subject.
Data Consolidation of Disparate Procurement Data Sources for Correlated Performance-Based Acquisition Nangia, Samantha; Dickover, Ryan; Wardwell, Thomas; Mora, Randall (Monterey, California. Naval Postgraduate School, 2017-03); SYM-AM-17-095Frank Kendall, then Under Secretary of Defense for Acquisition, Technology and Logistics, released the first defense acquisition system performance report in June 2013. This report focused primarily on performance related ...
Schaus, Brian M. (Monterey, California: Naval Postgraduate School, 2015-03);Techniques for classifying maritime domain targets-of-interest within images are explored in this thesis. Geometric and photometric features within each image are extracted from processed images and are used to train a ...
How does the cost performance of defense contracts vary among services and contractors? Evidence from major defense acquisition programs (MDAP) Euske, K.J.; San Miguel, Joseph; Wang, Chong (2012);This research examines how the cost performance of defense contracts varies among the Air Force, Army, Navy, and the Department of Defense (DoD) and among five major defense contractors: Boeing, Lockheed Martin, Northrop ...