Identifying the emergence of new technologies is a crucial task for most firms that strive to grow their businesses and stay competitive. We propose a quantitative forecasting approach that seeks to predict whether a technology will become emerging or non-emerging in one year, through the use of different machine learning algorithms. In order to determine the technologies that will comprise the foundation of our dataset, we used the Gartner Hype Cycle for Emerging Technologies framework from 2007- 2020. To develop a model, we rely on patent data which has been a stable source of information for measuring innovations and the characteristics of emerging technologies. By testing the models on unseen observations, we discovered that the most accurate algorithm was able to identify 67 pct. of all instances correctly. Through further investigation of performance and qualitative analysis with experts of strategic foresight, we conclude that companies should apply our proposed model as a monitoring tool and in conjunction with expert-centric approaches to anticipate technological advancements. Moreover, we suggest that the results are revisited and automated for improvements in the future.
|Educations||, (Graduate Programme) Final Thesis|
|Number of pages||172|
|Supervisors||Henrik Johannsen Duus|