Data Science

What is Data Science?

If you learn Bayesian statistics based on Bayes’ theorem, you can cultivate data literacy and you will be able to walk around the information-filled world without being confused. The Nakatsuma Study Group is also conducting research on AI and machine learning, and is applying advanced technologies using perceptrons, SVMs, and neural networks from code implementation to finance. By learning the basic technology of AI, you can build models and forecast markets that could not be realized by conventional statistical methods.


Quくん(Nakatsuma Seminer, 17th)

As a PEARL member of the Data Science part, I firmly believe that Nakatsuma Zemi is the most precious opportunity for me to fully transition my career path from Economics to Data Science and Artificial Intelligence. Among all zemis in Keio University, Nakatsuma Zemi is of the highest flexibility in my humble opinion. Zemi members do independent projects on topics of their own interests and present their progress in each regular seminar. On top of that, the Data Science part used to have a weekly sub-seminar, an intensive group study session of advanced topics in Data Science. Due to a lack of members last year, we did not hold this session weekly, but we still helped each other whenever we encounter troubles with theoretical understanding and Python implementation.
My project was to apply Bayesian methods (MCMC, Variational Inference, Gaussian Process) on top of machine learning models for post-COVID stock index forecasting. After evaluation with 2 different metrics, I found out the intrinsic advantage of pure Bayesian models in prediction tasks when facing extreme uncertainties such as the COVID-19 pandemic. This fruitful experience greatly enhanced my capabilities to conduct an AI project on my own.