Research Data Scientist (Wearable Accelerometry & Step Detection)

We are looking for a research-oriented data scientist to benchmark algorithms that convert raw 1-, 2-, and 3-axis accelerometer data into step counts. The project includes: (1) a literature review of industrial and scientific approaches and open datasets, (2) selection of suitable open-access accelerometry datasets, (3) Python implementation and parameter grid search of representative algorithms, (4) quantitative evaluation of accuracy versus reference step counts. The study should: - compare wrist- and hip-worn sensors, - assess the impact of reducing from 3-axis to 2-axis and 1-axis data, - summarize the trade-offs between accuracy, robustness, and computational complexity. Deliverables include well-documented Python code, a reproducible analysis notebook, publication-quality figures and tables, and a concise technical report suitable as the basis for a scientific manuscript. Time: 1 week