Data Scientist - Software Engineer - Ph.d. Student

Portfolio


My Research


Type Field Date Details
Journal Paper Machine Learning 2019 Hao Hua Sun Yin, Klaus Langenheldt, Mikkel Harlev, Raghava Rao Mukkamala, Ravi Vatrapu
"Regulating cryptocurrencies: a supervised machine learning approach to de-anonymizing the bitcoin blockchain"
Journal of Management Information Systems, Link to paper
Conference Paper Machine Learning 2018 Mikkel Alexander Harlev, Haohua Sun Yin, Klaus Christian Langenheldt, Raghava Mukkamala, Ravi Vatrapu
"Breaking bad: De-anonymising entity types on the bitcoin blockchain using supervised machine learning"
Proceedings of the 51st Hawaii International Conference on System Sciences (HICSS-51), Link to paper

Contact Me



Customer Lifetime Value Prediction


The CLV Prediction project is a joined initiative between CBS and Danske Bank. Danske Bank seeks to be more relevant to their customers by predicting their future profitability to the bank, a common metric by the name of Customer Lifetime Value (CLV). By understanding future profitability the bank can allocate adequate financial resources to prevent unnecessary risk (e.g. default) leading to increased customer satisfaction as well as increased profit. To predict such an error-prone target I am utilising recurrent neural networks in general and LSTM in particular, as well as Markov Modelling.

Close Project

Blockchain Forensics


The Blockchain Forensics project is a joined initiative between CBS and Chainalysis, market leader in Bitcoin blockchain forensics. Me and my two colleagues, Klaus Langenheldt and Awa Sun Yin crunched proprietary labelled Bitcoin Blockchain data to uncover types of actors (referred to as entity types) on the Bitcoin Blockchain. Our Gradient Boosting model, built in Python, was able to classify all bitcoin addresses on the Bitcoin Blockchain. Read more in our papers from HICSS and JMIS.

Close Project

Kredsløb


Kredsløb is a start-up launched in 2019. Our primary goal is to make it easier for everyone to contribute to sustainability and save money on their electricity arrangements. We automatically scan the energy market to find the most suitable electricity arrangement for our members.

Check our website! Close Project

T2S Project


For a couple of years I worked on the T2S project as a software developer and on-call specialist. T2S is a EU initative with the purpose of easing security trade across borders. I was primary developing mainframe applications supporting all our mortgage loan applications ammounting to several hundred billions DKK annually.

Close Project