For many start-ups, lack of investment and capital has become the bottleneck for development. This phenomenon inspires us to use machine learning algorithms to find patterns in investment behavior from major investors. We plan to use various domain-specific features to predict which investors would potentially invest in a particular company . This would not only reveal important information about investment strategies and behaviors of investors, but also give startups ideas on where to seek potential investment and how to adjust their strategies so as to attract potential investors.
Our work is grounded in CrunchBase, an accessible knowledge base that maintains full records of company and people information.
There are two primary goals of our work:
- To predict whether an investor would invest in a particular start-up based on textual, topological and domain-specific signals from both the investor and start-up.
- To analyze and reveal the factors that would prompt an investor to invest in startups so as to shed light on the adjustments the start-ups could make to attract more investments.