Projekt

Olga Otinova - Data Science Semesterprojekt

Development of a Cost-Effective Production Portfolio Model of Innovation Projects

9 out of 10 start-ups do not come out of the "valley of death", this is a period of high uncertainty when you have to look for answers to the questions of whether the start-up is developing properly, whether it will turn out to be a real business. There are many reasons that influence and hinder a startup, such as an unsuccessful business model, high competition, and others.
The hardest thing, in that case, is to answer the question if the startup is successful and predict its future. Only the most experienced experts are capable of this. And whether it is possible to predict success based on the experience of other companies, and which parameters are more important for evaluating a startup, I tried to answer these and other questions during the project.
As a result of the project, it was possible to obtain a classifier that predicts, as a percentage, how successful the company will be according to its parameters. The criteria that have the greatest influence on the success of the project are also highlighted. However, the project does not end there, the next steps in its development are as follows:

- Evaluate success and compare companies in a category. For example in type of industry, or similar size of investment;
- Set up clustering and to try other approaches. Find interesting patterns;
- Add the ability to rate a company from outside the dataset.