Strategic analysis of Influence Maximization

A large part of technological innovation in fast-changing industries is related to start-up companies who make an important contribution to the UK's competitive position in emerging industries. However, chances of success of start-up companies are currently only poorly understood and factors related to the availability of expertise and knowledge via social networks have not found much attention in the literature so far. Further insights about IPO success are potentially useful for (i) policy makers adjusting industrial policy to help innovation, (ii) financial services companies and banks deciding about loans for start-up companies, and (iii) early stage investors such as Angel investors.

This PhD project aims at addressing this gap in knowledge and will apply techniques from machine learning and network science to available big data bases which combine information about start-up companies, their employees, funding and ultimate IPO success, to gain insights into factors that influence start-up success. The project will build on recent work by Dr Panzarasa (to join the supervision team as an external supervisor) who demonstrated that centrality in an employee-company network influences success. More precisely, the project aims to achieve the following goals:

1st year: Construction and statistical characterisation and analysis of big employer-company-director-funder networks gathered from online sources.

2nd year: Statistical characterisation of factors contributing to company success. Application of machine learning techniques and feature engineering to develop a model of company success.

3rd year: Development of statistical models of network formation and analysis of the underlying dynamical processes