Organisers: Miguel Delgado,
Javier Hidalgo and
Oliver Linton
Assistant: Sue Kirkbride (
s.kirkbride@lse.ac.uk, 020 7955 7509)
Date: Friday 25th - Saturday 26th May 2007
Venue:
Michio Morishima Room (R505) 5th Floor,
LSE Research Laboratory
London School of Economics and Political Science,
Houghton Street,
London WC2A 2AE
View Program >>
Peter Robinson was born in England in 1947. In 1968 he received his degree in Statistics at University College London, and he completed the M.Sc. in this discipline at the London School of Economics, where he served as Lecturer in 1970. In this same discipline he obtained his PhD at the Australian National University in 1973 under the supervision of Prof. Edward
Hannan. After working at the universities of Harvard, British Columbia and Surrey, he returned to the London School of Economics in 1984, where he currently occupies the Tooke Chair in Economic Science and Statistics, the highest scientific distinction of this institution. The London School of Economics has lead for decades the implementation of Statistics in Social Sciences, particularly Economics, developing probabilistic models able to explain economic agents behaviour with the help of Economic Theory; that is, Econometrics. Professor Robinson has contributed significantly to the impressive advance of this discipline in the last decades, for which he has been awarded the titles of Fellow of the British Academy, the Econometric Society and the Institute of Mathematical Statistics. He has supervised 23 doctoral researchers.
If you would like to attend, contact
Javier Hidalgo
Confirmed Speakers:
You may
download the Programme (in Adobe PDF format) or view it below:
DAY ONE - Friday 25th May 2007
|
| 9:15 - 9:30 |
OPENING AND WELCOME
Oliver Linton and Javier Hidalgo
|
|
SESSION 1 (Chair: Ignacio Lobato) |
| 9:30 - 10:20 | Cheng Hsiao
Parametric, Semi-parametric and Nonparametric Assessment of Mother's Working Status on Childhood Obesity |
| 10:30 - 11:20 | Whitney Newey
Choosing the number of Moments in Conditional Moment Restriction Models |
| 11:30 - 12:00 | Coffee Break |
| 12:00 - 12:50 | Andrew Chesher
Partial Identification with Discrete Outcomes |
| 13:00 - 14:30 | LUNCH
|
| SESSION 2 (Chair: Domenico Marinucci) |
| 14:30 - 15:20 | Murad Taqqu
Self-Similar Stable Processes and Flows |
| 15:30 - 16:20 | Rainer Dahlhaus
Statistical Inference for Locally Stationary Processes |
| 16:30 - 17:00 | Coffee Break |
| 17:00 - 17:50 | Hans Künsch
A Random Walk Through 25 Years of Time Series Analysis
|
| DINNER
|
| 18:00 - 19:00 | Reception in the Senior Common Room |
| 19:00 - 22:00 | Dinner in the Senior Dining Room
|
DAY TWO - Saturday 26th May 2007 |
| SESSION 1 (Chair: Paolo Zaffaroni) |
| 9:30 - 10:20 | Peter Phillips
Long Memory and Long Run Variation |
| 10:30 - 11:20 |
Peter Hall
Bootstrap Aggregation for Cross-Validation and Inference Under Constraints
Abstract:
A major asset of the cross-validation approach to smoothing-parameter choice is its utilitarian character. However, the bandwidths produced by cross-valdiation are relatively noisy, and this difficulty impedes good performance. The stochastic variability of cross-validation can be reduced significantly by using bootstrap aggregation, or bagging, a method proposed by Breiman (1999). The technique is very simple to use, and enjoys the utilitarian character of cross-validation. For instance, it can be applied in practically all of the many settings where cross-validation is employed for bandwidth choice with the aim of optimising an $L_2$ measure of performance. Arbitrarily large reductions in bandwidth variability are theoretically possible, although in practice bagging would likely be used relatively modestly. In particular, half-sample bagging can reduce bandwidth variability by approximately 50%. We shall also discuss a bagging-based approach to constrained inference, for example to parameter estimation when it is known that the true value of the parameter satisfies an inequality constraint.
|
| 11:30 - 12:00 | Coffee Break |
| 12:00 - 12:50 | Yuzo Hosoya
Inference On Transformed Stationary Time Series
Abstract:
The paper presents an approach to deal with parametric inference on instantaneously transformed stationary processes, showing that the information equality and hence the conventional large-sample test theory do not hold in general for large-sample parametric inference on such processes. The paper newly introduces a modified version of the Box-Cox transformation and as a specific case investigates in detail that version of Box-Cox transformed vector ARMA processes, proposing a computation-intensive procedure for parametric estimation and testing. Illustrative simulation and real-data examples are provided.
|
| 13:00 - 14:30 | LUNCH
|
| SESSION 2 (Chair: Carlos Velasco) |
| 14:30 - 15:20 | Donald Andrews
Hybrid and Size-Corrected Subsample Methods |
| 15:30 - 16:20 | Xiaohong Chen
Efficient Estimation of Sequential Moment Restrictions Containing Unknown Functions
Abstract:
This paper studies semiparametric efficient variance bounds and efficient estimation of smooth functionals for models of sequential moment restrictions containing unknown functions, in which some conditioning variables for one moment equation could be endogenous variables to another moment equation, and some of the unknown functions could be functions of endogenous variables. The class of models extends those of Newey and Powell (2003) and Ai and Chen (2003) to allow for conditional moment restrictions with different conditioning variables. The semiparametric efficiency bound results extend those of Chamberlain (1992), Hahn (1997) and Brown and Newey (1998) to allow for moment restrictions involving unknown functions of interests. As a non-trivial example, we provide semiparametric efficient estimation of the weighted average derivative in a nonparametric IV regression model.
|
| 16:00 | Peter Robinson |
.
