With every deed you are sowing a seed, though the harvest you may not see.

In late October, I attended the 2022 Technology Hackathon held by Natixis. The event was promoted by CUHK, and the task given was to detect fraudulent Ethereum addresses by applying machine learning models to Ethereum transactions. 

I’ve picked up quite many data science skills during the four years of undergraduate studies, so when I found out the challenge was closely related to my major and interest, I registered for this hackathon as a trial. The result was kind of unexpected for me, because the major goal for me was only to gain more real-life experience.

The whole process of the challenge combined both the business perspective (e.g. how to build meaningful features and how to present the model) and the technology perspective (e.g. how to build the model and how to interpret the data distribution). As a result, the task was a good practice for those who would like to familiarise themselves with the real-life production process. In terms of the dataset, participants were given a dataset of normal and fraudulent addresses, along with their account activities, such as the number of sent transactions, received transactions and transferred ETHs.

As I became more familiar with supervised learning tasks, I treated the task I was given as a traditional classification task (using algorithms like Boosting Classifier, Logistic Regression Classifier), and constructed features of statistical meaning for each account using the transaction history.

The result is more than a big encouragement for me to continue my path in the field of data science and fintech. I am glad to have the chance to practise, and to explore model and visualisation techniques out of my original knowledge scope.

Words of Encouragement:

When I decided to pursue my future path in data science three years ago, I did not even know what data science exactly was. As such, I began my long way of exploration in this field. What drives me continuously has been the abstruse interpretation in the world of data, the process to continuously learn new knowledge about machine learning and many specific areas. 

I believe that with proper use of the tools we have, life could be made better. And data science is just one way to connect the dots of technology and our daily lives, in the fields of finance, healthcare, human language interpretation, and more.

I started my first prediction project on my data mining course two years ago. Since then, I’ve been actively participating in different projects. I am glad that I can gradually build my skillsets, but I know there is still a long way for me to go. I would like to hereby thank all my friends and professors who always encourage me.

Lastly, I would like to share with you one saying that has been a source of inspiration for me since high school. That is, “with every deed you are sowing a seed, though the harvest you may not see”. I hope we all can one day see the harvest.