Prediction of capital issues of Investment Banks

OData support
Supervisor:
Dr. Kósa Zsuzsanna Mária
Department of Telecommunications and Media Informatics

The goal of the thesis is to create an algorithm and a system that can analyze data of firms that listed on stock exchange and predict the time of a firm next going to the capital market.

The algorithm first cleans the data than it uses prediction algorithms to see when will a company next try to go to the market by the help of investment banks.

For this first, I give information about why does a company needs money and what kind of methods can a company use to get financing, how does the stock exchange works. It gives brief knowledge about how does the debt capital market works. The data that I work on needs analysis and a lot of cleaning.

With the gathered business knowledge and cleaned data, I use different algorithms to predict the next issue date.

The system created in Microsoft Azure. For cleaning the data besides Microsoft Azure, I also create R code and use Microsoft SQL Studio. To find the best fitting algorithm in the thesis I try several prediction algorithms and compare the result of these to find the best one.

Downloads

Please sign in to download the files of this thesis.