Comparison of Machine Learning Models for Predicting the Impact of Acquisitions on Innovation in the Defense Sector

With the Department of Defense identifying innovation as a key enabler for maintaining military superiority, the defense industry has made increasing this capability a priority. Defense companies have participated in mergers and acquisitions to augment internal research and development to meet the Department of Defense’s goals. Selecting acquisitions that will increase innovation is a challenging task for defense companies as wrong acquisitions can decrease the acquiring company’s ability to innovate. This praxis employed the patent success ratio as a metric for measuring innovation, and then used different machine learning techniques to predict changes to Patent Success Ratio resulting from acquisitions at 3- and 5-year periods after acquisition. Those models were then compared across several metrics. The research resulted in models that were able to predict with more than 75% accuracy the effect on Patent Success Ratio for the 3-year period after acquisition and over 80% for and 5-year periods post-acquisition performance. These tools will enable decision makers in the defense industry to make informed decisions when selecting acquisitions to increase innovation. Additionally, the praxis shows that acquisition count alone is not an indicator of increased innovation. Results show that ensemble models using several different techniques provide the best prediction of innovation improvement for defense companies.

Author

• Focacci, Aaron

Language

• en

Keyword

• Innovation
• Crunchbase
• Machine Learning

Date created

• 2022

Type of Work

• Dissertation

License

• All rights reserved

Rights statement

• In Copyright

GW Unit

• Engineering Management

Degree

• D.Engr.

Advisor

• Yassan, Rebecca

Committee Member(s)

• Dimitriou, Ari
• Campos-Nanez, Enrique

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/73666543r