Bulletin Number One 1983

The New Depa r tment of Statistics As the social needs for Statistics in various fields, such as commerce, industry and education, steadily grow, the University finds it essential that the Statistics Section, which was set up in 1978-79 under the Department of Mathematics, be developed into a full-fledged Department. In the current academic year, an independent Department of Statistics was established ill the Faculty of Science, offering the following courses: (1st year) Pro-Seminar, (2nd year) Fundamentals of Statistics, Statistical Programming, Introductory Statistics I & II, (3rd year) Experimental Designs, Applied Regression, Statistical Quality Control, Acturarial Science, Survey Sampling, Data Analysis, (4th year) Applied Multivariate Analysis, Time Series and Forecasting, Computational Statistics, and Path Analysis to its major and minor students. This year, the number of major students is seventeen, which is the quota for majors. There is, however, no fixed quota for minors and the enrolment comes to 327. As the programme continues to consolidate and develop, it is anticipated that enrolment will steadily increase. A graduate programme in Applied Mathematics including Statistics was offered by the Mathematics Division of the Graduate School before the establishment of the Division of Statistics in 1981. This graduate Division now offers an M.Phil. Degree Programme in Statistics. This year, the number of graduate students totals five. In order to promote academic exchanges and research interests in Statistics, the Statistics Department holds regular seminars. Each member of the Department has his own research interests: Dr. N.N. Chan — regression models and regression designs, and functional and structural relationships; Dr. Lee Sik Yum and Dr. Leung Chi Ying - methods in multivariate analysis; Dr. Lam Hing Kam — the calculation problems in statistical distributions. The research interests of Professor Howell Tong can be seen in his profile. The University is considering the introduction of a part-time undergraduate programme in Statistics in 1987-88 and is exploring the possibility of offering a Part-time Master or Diploma Course in Managerial and Social Statistics under the Department. Professor Howe ll Tong Professor of Statistics Professor Howell Tong was born in 1944 in Hong Kong. He was awarded his B.Sc. degree (with first class honours in Mathematics) in 1966 and Ph.D. degree in 1972 at the Victoria University of Manchester. He was appointed a demonstrator at his alma mater in 1968 and reached the rank of Senior Lecturer in 1977 He joined this University in August 1982 to take up me newly created Chair of Statistics in the Science Faculty . Professor Tong is a Fellow of the Royal Statistical Society. In 1974, he was sponsored by the Royal Society of London as a Visiting Fellow to Japan and spent six months at the Institute of Statistical Mathematic Tokyo. In 1981, he conducted a two-month lecturing tour of China, at the invitation of Professor L.K. Hua of the Institute of Applied Mathematics (Beijing), again supported by the Royal Society of London. He has published about forty papers in reputable journals such as the Journal of Royal Statistical Society, Biometrika ， Journal of Applied Probability ， Transactions of Institute of Electrical and Electronic Engineering, International Journal of Control and Journal of Time Series Analysis. His speciality is time series analysis, a subject to which he has devoted over fifteen years. He is particularly interested in extending existing methodology of time series analysis to non-stationary and nonlinear situations of important practical significance. The contents of his doctoral thesis formed the basis of a paper on non-stationary time series, which he read jointly with his supervisor, Professor M.B. Priestley, to the Royal Statistical Society in 1972. In 1980’ he read a paper on non-linear time series to the same learned Society. Besides, he has a very genuine interest in practical applications, covering forecasting, automatic control, information science, signal processing, commerce, ecology, hydrology, geophysics, meteorology' etc. In the last five years, he has developed a new methodology of non-linear time series analysis, generically called the threshold models, and is completing a monograph on the subject. 16 RECENT DEVELOPMENTS