Vahid Farrahi is a doctoral researcher in Medical Technology at the Faculty of Medicine, University of Oulu, Finland, since 2017. He received his MSc in Information Technology from the Faculty of Computer science and Information Technology, Shiraz University of Technolgy, Iran in 2016. Over the past 5 years, he has been working with novel data analytical approaches including data mining and machine learning methods to tackle real-world problems. His research interests include data mining, machine learning, signal processing, and biomechanics. Vahid's PhD project is currently funded by BioMEP (https://www3.uef.fi/web/biomep) under the Marie Skłodowska-Curie Actions (MCSA) grant agreement No 713645.
- Data Analytics
- Machine learning
- Data mining
- Health technology
- Signal processing
PhD. project in brief
Physical activity is well-known by general public to be a necessity for maintaining health and well-being, which is completely true. To date, it has been proved that physical activity highly correlates with certain type of health problems including diabetes and cancer. For having a better physically active society, providing recommendations at individual and public level regarding physical activity, and getting insights about physical activity and sedentary behaviors in a society, estimation of physical activities in terms of duration, intensities, and energy expenditure on physical activities are necessary non-trivial tasks.
Objective measurement of physical activity data using accelerometers is a practical method of assessing physical activity behaviors. However, it is still a challenge to accurately estimate the amount, intensity and type of physical activities using accelerometer data. To date, (traditional) statistical methods were applied on physical activity data to assess physical activity behaviors. The overall goal of the project is the transition from traditional statistical methods to (big) data mining methods to assess their accuracy and applicability, in order to get better insights about physical activity and sedentary behaviors. Furthermore, as the project constituents follow-up data (including biological samples and subjects’ characteristics), relating various health variables to physical activity behaviors through big data analytics is of great importance as a goal. Ultimately, the health and well-being will be promoted by providing more accurate insights for health practitioners.
Professional and community activities
- IEEE member
- Reviewers for Gait & Posture, Scientific Reports, and Systematic Reviews.
- Doctoral student, Research Unit of Medical Imaging Physics and Technology
- Farrahi, Vahid; Niemelä, Maisa; Tjurin, Petra; Kangas, Maarit; Korpelainen, Raija; Jämsä, Timo (2020) Evaluating and Enhancing the Generalization Performance of Machine Learning Models for Physical Activity Intensity Prediction from Raw Acceleration Data. - IEEE journal of biomedical and health informatics 24 (1), 27-38 . [Original] [Self-archived]
- Farrahi, Vahid; Niemelä, Maisa; Kangas, Maarit; Korpelainen, Raija; Jämsä, Timo (2019) Calibration and validation of accelerometer-based activity monitors: A systematic review of machine-learning approaches. - Gait and posture 68, 285-299 . [Original] [Self-archived]
- Niemelä, Maisa; Kangas, Maarit; Farrahi, Vahid; Kiviniemi, Antti; Leinonen, Anna-Maiju; Ahola, Riikka; Puukka, Katri; Auvinen, Juha; Korpelainen, Raija; Jämsä, Timo (2019) Intensity and temporal patterns of physical activity and cardiovascular disease risk in midlife. - Preventive medicine 124, 33-41 . [Original] [Self-archived]
- Niemelä, Maisa; Kiviniemi, Antti; Kangas, Maarit; Farrahi, Vahid; Leinonen, Anna‐Maiju; Ahola, Riikka; Tammelin, Tuija; Puukka, Katri; Auvinen, Juha; Korpelainen, Raija; Jämsä, Timo (2019) Prolonged bouts of sedentary time and cardiac autonomic function in midlife. - Translational Sports Medicine 2, 341–350 . [Original] [Self-archived]
Bradford Research Institute for Health (BRIH), Bradford, UK
A 6-month placement to perform novel analytical approaches (compositional data analysis) to investigate the associations between objectively measured physical activity and cardiometabolic outcomes in a large-scale population-based cohort (Sept 2019-March 2020).