Driven By Data.
Insights.
Curiosity.
This is the personal website of Mai Tanaka, PhD, a data-oriented researcher specializing in brain signal analysis.
Here you’ll find my projects that reflect my interests—including modeling my running performance from training and physiological data—along with technical tutorials and practical insights.
Here are some of the things I can do:
Featured Projects
From brain waves to fitness data, here are my current favorites:
Marathon Training Analysis
Train Light. Perform Hard.
Predicting running pace using recent training load, heart rate data, and environmental conditions.
This project explores how training load, physiological signals, and environmental factors influence running performance. I combined Fitbit exercise data with weather data from the Japan Meteorological Agency, building a data pipeline to align and clean multiple real-world data sources. This included filtering and structuring time-series workout data, correcting time zone inconsistencies, and integrating temperature and humidity conditions for each run. I then applied statistical modeling and feature engineering to predict running pace for higher-effort runs based on recent low-intensity training, heart rate, and environmental variables. The project demonstrates end-to-end data analysis, from data collection and preprocessing to modeling and interpretation in a real-world setting.
Decoding Imagined Speech
Under construction…
Robot Arm Calligraphy
Under construction…
Whale Song Translation
Under construction…
Tips, Tricks and Thoughts
Blog posts on technical tutorials, useful tools and some cool research.
Data Scraping from JMA
Scrape tabular weather data from the Japan Meteorological Agency’s website with this step-by-step technical tutorial
Exploring Fitbit Exercise Log
Click here to change this text. Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
About Me
Applying data science to uncover insights in biological data.
I’m a data-oriented researcher with postdoctoral experience working with high-dimensional time-series data, including EEG and MEG. I began my work in MATLAB; however, I now primarily use Python for preprocessing, analysis, statistical modeling, and signal processing. My work focuses on turning complex, noisy data into clear and actionable insights.
My research background has shaped a structured, hypothesis-driven approach to problem solving. As a result, I work effectively with messy, real-world data. In addition, I prioritize clarity, reproducibility, and rigor in everything I build, from data pipelines to analytical models.
I approach problems systematically and break them into smaller, testable components. At the same time, I communicate results clearly and accessibly across both technical and non-technical audiences.
Currently, I am building practical data science projects and exploring applications in health, fitness, and medicine. Through this work, I continue to strengthen my skills in applied data analysis and real-world problem solving. This site showcases my projects, technical tutorials, and reflections on productivity and analytical thinking.
Outside of data science, I enjoy reading Japanese novels, playing soccer and solo traveling.
If you’re working with brain signal data or building tools in this space, I’d be glad to connect.