Pekka Siirtola

Pekka Siirtola

Dr. (Tech)

Postdoc researcher
Machine learning, Activity recognition, Wearable sensors, Data mining

39 peer reviewed publications in international journals. The publications have been cited 382 times, of which the most cited publication accounts for 111 citations. The author has a h-index of 8.00. (Google Scholar)

Biography

I received my Ph.D. in 2015 and since that I have worked at the DataAI group, a sub-group of Biomimetics and Intelligent Systems Group, as a postdoctoral researcher. Most of my publications are related to analyzing data from wearable and mobile phone sensors, and especially, using this data to recognize human activities.

Research interests

  • Machine learning
  • Activity recognition
  • Wearable sensors
  • Data mining

Research topics

  • Activity recognition

I have a long history in studying human activity recognition based on inertial sensor data. Mainly I have focused on studying accelerometer data from smartphone and wearable sensors. Application areas have varied from detection of daily activities, and swimming styles to detecting industrial activities. Recently study has focused on training adaptive models and building general models based on small training data.

  • Migraine attack detection

One of the recent research topics is early detection of migraine attack based on wearable sensor data. Idea is to use wearable sensors to measure biosignals like, blood volume pulse, heart rate variability and galvanic skin response, and find out how these change before forthcoming seizure or attack.

  • Continuous authentication

The aim of the continuous authentication is to find out if smartphone is used by genuine user or impostor. This authentication is based on touchscreen and other smartphone sensor data collected during user's swipe gestures.

Research groups

  • Postdoctoral researcher, Biomimetics and Intelligent Systems

Selected publications

  • Siirtola, Pekka; Koskimäki, Heli; Röning, Juha (2016) Personal models for eHealth - improving user-dependent human activity recognition models using noise injection. (Artikkeli tieteellisessä konferenssijulkaisussa). - 2016 IEEE Symposium Series on Computational Intelligence (SSCI). Ei sarjaa/No series. 1-7. [Original] [Self-archived]
  • Siirtola, Pekka (2015) Recognizing human activities based on wearable inertial measurements : methods and applications. - Acta Universitatis Ouluensis. Series C, Technica 524. Oulu. Artikkeliväitöskirja. 94, [56]. [Original]
  • Siirtola, Pekka; Röning, Juha (2013) Ready-to-use activity recognition for smartphones. (Artikkeli tieteellisessä konferenssijulkaisussa). - 2013 IEEE Symposium on Computational Intelligence and Data Mining (CIDM). IEEE, IEEE. 59-64
  • Pekka Siirtola; Juha Röning (2012) Recognizing Human Activities User-independently on Smartphones Based on Accelerometer Data. - International Journal of Interactive Multimedia and Artificial Intelligence 1 (5), 38-45 . [Original]
  • Pekka Siirtola, Heli Koskimäki, Ville Huikari, Perttu Laurinen, Juha Röning (2011) Improving the classification accuracy of streaming data using SAX similarity features. - Pattern Recognition Letters 32 (13), 1659-1668

Research visits

  • Cross-sectoral research visit

    3/2017 - 9/2017: Research visit to Bittium Wireless Ltd