s
  • When the Department was founded, the teaching staff consisted only of a professor and an assistant. 
  • At present there are 9 teaching positions: two professors, three lecturers, senior assistant, assistant and two full-time teachers. There are also a number of research staff on fixed-term contracts.
  • Also a large number of Statistics part-time teachers have worked in the Department. These include Tomi Seppälä, Anna-Maija Koivisto,Ismo Lapinleimu, Henri Toukomaa, Merja Jauhiainen, Markopekka Niinimäki, Joni Lehtinen, Matti Ylén, Jari Parviainen, Pirjo Palmroos, Inka Martti, Lasse Koskinen, Hannu Oja, Jani-Pekka Virtanen, Minna Åhman and Erkki Mäkelä.
    		•

    Pentti Huuhtanen

    Ph.D., Lecturer

    E-mail: pvh@uta.fi

    Pentti Huuhtanen obtained his M.Sc. (1964) at the University of Helsinki, and Lic. Phil. (1977) as well as Ph.D. (1998) at the University of Tampere. He was appointed lecturer (1970) in the University of Tampere and has held the position since that time. He has been an acting Professor in Statistics (1985–86, 1987, 1994) and a senior assistant (1997). Outside the academic world he has experience of work in the electric power industry and school teaching.
     

    Research interests:

    • time series analysis
    • stochastic complexity
    • neural networks.

    Recent publications

    • Lehtokangas, M., Saarinen, J., Huuhtanen, P. and Kaski, K. (1993). Neural Network Prediction of Non-Linear Time Series Using Predictive MDL Principle, IEEE Winter Workshop on Non-Linear Digital Signal Processing (Jan 17-20 1993, Tampere Finland)
    • Lehtokangas, M., Saarinen, J., Huuhtanen, P. and Kaski, K. (1993). Neural Network Modeling and Prediction of Multivariate Time Series Using Predictive MDL Principle, ICANN'93 Proceedings of the International Conference on Artifical Neural Networks (Sept 13-16 1993, Amsterdam The Netherlands)
    • Lehtokangas, M., Saarinen, J., Huuhtanen, P. and Kaski, K. (1993). Chaotic Time Series Modeling with Optimum Neural Network Architecture, Vol. 3 of 3 Proceedings, International Joint Conference on Neural Network IJCNN'93 (Oct 25-29 1993, Nagoya Japan)
    • Lehtokangas, M., Saarinen, J., Huuhtanen, P. and Kaski, K. (1993). Neural Network Optimization Tool Based on Predictive MDL Principle for Time Series Prediction, Proc. of the 1993 IEEE Int,'1 Conf. on Tools with AI (Nov 1993, Boston Massachusetts)
    • Lehtokangas, M., Saarinen, J., Huuhtanen, P. and Kaski, K. (1993). Time Series Modeling Using Predictive MDL Principle for Model Selection, Submitted for Publication
    • Lehtokangas, M., Saarinen, P., Huuhtanen, P. and Kaski, K. (1995). Initializing Weights of a Multilayer Perceptron Network by Using the Orthogonal Least Squares Algorithm. Neural Computation, vol. 7, no. 5, pp. 982-999
    • Lehtokangas, M., Saarinen, J., Huuhtanen, P. and Kaski, K. (1995). A Network of Autoregressive Processing Units for Time Series Modeling. Applied Mathematics and Computation 75: 151-165
    • Lehtokangas, M., Saarinen, J., Huuhtanen, P. and Kaski, K. (1995). Predictive Minimum Description Length Criterion for Time Series Modeling with Neural Networks. Neural Computation 8, 583-593
    • Huuhtanen, Pentti (1998). Stochastic Complexity and the MDL Principle in Density Function Estimation. Acta Universitatis Tamperensis 611, University of Tampere, Tampere, Finland. (Dissertation)

    << Back <<