Skip to content. | Skip to navigation

Informatik 5
Information Systems
Prof. Dr. M. Jarke
Personal tools
You are here: Home Theses A machine learning approach to investigate startup success


Prof. Dr. M. Jarke
RWTH Aachen
Informatik 5
Ahornstr. 55
D-52056 Aachen
Tel +49/241/8021501
Fax +49/241/8022321

How to find us

Annual Reports





A machine learning approach to investigate startup success

Thesis type
  • Bachelor
Status Running
In this project a student has to apply machine learning algorithms to new venture foundations, to learn about inherent process sequences that successful and unsuccessful founders engage in. We provide the student with the data on founder characteristics, gestation activities and venture characteristics (Reynolds & Curtin, 2011). It longitudinally follows founders over five subsequent annual waves (2007-2011). The data includes facts about founders' activities such as opening or closing a new office, selling a new product, etc.

Methods We want to apply a recent development in deep learning: a Recurrent Neural Network(RNN) that respects input of sequential data (time-series). The networks can be realized with the help of TensorFlow (Abadi et al. 2015) or other frameworks. Input data includes all activities of entrepreneurs over 192 months that we use to predict entrepreneurial action sequences that lead to successful ventures. The results can be evaluated with sequential pattern mining approaches.

Result Overall, such an analysis might help us identifying the sequences that determine the best path to startup success.
Document Actions