graphs/graph_enac_l.gif
English only
INTER  >  TRANSP-OR > Workshop on Discrete Choice Models
Logo TRANSP-OR
Content

Workshop on Discrete Choice Models

August 19 - 21, 2010

Ecole Polytechnique Fédérale de Lausanne, Room GC B3 30

The 2010 workshop will be organized in the same spirit as the previous ones: an informal meeting for the exchange of ideas around discrete choice models, with the objective to trigger new collaborations, or strengthen existing ones. At the end of the workshop, a list of potential collaborations will be identified, with specific objectives.

Registration

Top

Everyone interested is invited to attend. Presentations are upon invitation only. All participants, including speakers, must register with the following form.

The registration fee (CH 130) includes the dinner on Thursday, lunch on Friday, coffee breaks, as well as transportation on Saturday.

Click here to access to the registration form

Schedule

Top
Thursday Aug. 19, afternoon - Friday Aug. 20, morning
Presentations (program not yet available)
Friday Aug 20, afternoon
Workshop meeting
Saturday August 21, morning
Hiking in the vineyards
Saturday August 21, 12:00
Genuine swiss fondue in Vevey
Thursday
14:00 14:15 Welcome
14:15 14:45 Ben-Elia Modelling the effect of information accuracy and compliance on route-choice behaviour
14:45 15:15 Axhausen Estimating schedule choice models: First experiences
15:15 15:45 Bierlaire Modeling Route Choice Behavior From Smartphone GPS data
15:45 16:15 Break
16:15 16:45 Glérum Analysis of the impact of travelers' attitudes and perceptions on mode choice in low-density areas
16:45 17:15 Cherchi Accounting for inertia in modal choices: Some new evidence using RP/SP dataset
17:15 17:45 Robin Modeling the behavior of investors



Friday
09:00 09:30 Ramjerdi Application of cumulative prospect theory for the estimation of reliability
09:30 10:00 Flötteröd Metropolis-Hastings sampling of alternatives for route choice models
10:00 10:30 Knockaert The Spitsmijden experiments: the impact of rewards on departure time choice
10:30 11:00 Break
11:00 11:30 Chen Modeling the choice behavior underlying the usage of smartphones
11:30 12:00 Hurtubia Bootstrapping approach for sampling of alternatives in MEV models

Accomodation

Top

EPFL has negociated special rates with several hotels in the area.

The complete list is available here.

Ask the hotel for a free public transportation pass.

Transportation

Top

The easiest way to get to EPFL is to take the train from Geneva Airport to Renens. In Renens, take the light-rail (called M1) towards Lausanne. There is a stop at EPFL. The travel time is about 1 hour.

A map of the bus and metro network can be found here and time tables are available at the Lausanne Transport web page. Ask the hotel for a free public transportation pass.

Check the Swiss Federal Railways website.

To navigate within EPFL, use map.epfl.ch.

Consult also the page "How to get to EPFL?"

List of presentations (11)

Top

Click on the title to download the slides (if available).

List of abstracts

Top
  1. Estimating schedule choice models: First experiences by Kay Axhausen (ETH Zürich)

    The choice models estimated and used for transport planning stylize the choice situation by tradition and design. This is even more pronounced as the number of choice dimensions is increased in nested logit or similar modelling forms. The presentation will reflect the issues arising from this disconnect using the results of a recently concluded PhD dissertation. In the thesis a MNL of the choice between complete daily schedules is estimated and tested through simulation. The necessary hand-calibration is the starting point of the talk.

    The presentation will briefly describe the results of the thesis, discuss this experience and will then generalise it by including the issues raised through the disconnect mentioned above.

    (with M. Feil)

  2. Modelling the effect of information accuracy and compliance on route-choice behaviour by Eran Ben-Elia (University of West of England)

    ATIS (Advanced traveller information systems) are designed to assist travellers in choice making. Choice behaviour and especially compliance with the system's recommendations, is likely to be influenced by the system's level of accuracy (i.e. the difference between expected and actual travel times). Nested logit and mixed logit models are presented based on SP data obtained from a repeated route-choice experiment on a 3-route network with feedback. The results suggest behaviour is sensitive to accuracy levels and is also dependent on compliancy which can be best modelled as a nest parameter.

  3. Modeling Route Choice Behavior From Smartphone GPS data by Michel Bierlaire (EPFL)

    Smartphones have the capability of recording various kinds of data from built-in sensors such as GPS in a non-intrusive, systematic way. In order to be used as observations for route choice models, the discrete sequences of GPS data need to be associated with the transportation network to generate meaningful paths. In this paper, a probabilistic path generation algorithm is proposed to replace conventional map matching (MM) algorithms. Instead of giving a unique matching result, the proposed algorithm generates a set of potential true paths, along with probabilities for each one to actually be the true path. Temporal information (speed and time) is used to calculate the probability for observing the data while traveling on a given path. Comparisons against a state of the art deterministic MM algorithm using real trips recorded from a single user's smartphone are performed so as to illustrate the robustness and effectiveness of the proposed algorithm. Also, a Path-Size Logit (PSL) model is estimated based on a sample of real observations. The estimation results shows the viability of applying the proposed method in a real context.

  4. Modeling the choice behavior underlying the usage of smartphones by Jingmin Chen (EPFL)

    The data collected by smartphones contains rich information about the user's context and activities. However, the utilization of this data to infer the activities of a user is not an easy task. The frequently proposed machine learning methods lack an underlying behavioral model and are therefore unable to extrapolate beyond the data set based on which they were trained. We propose to model phone usage as a sequence of choices in different activity situations, and to use this model to infer the activites of a user from her phone usage.

  5. Accounting for inertia in modal choices: Some new evidence using RP/SP dataset by Elisabetta Cherchi (University of Cagliari)

    Inertia measures the effect that experiences in previous periods have on the current choice. As such it measures the tendency of sticking with the past choice or the disposition to change, when some alternative becomes particularly appealing. At the same time new situations force individuals to rethink about their choice and new preferences are formed. A learning process begins that release the effect of inertia in the current choice. Using a mixed dataset of revealed preference (RP)-stated preference (SP) we test and compare several ways of measuring inertia. We explore some new measures of inertia to test if inertia is stable along the SP experiments and we disentangle this later effect from the pure inertia effect between RP and SP. Finally we explore the relation between the utility specification (especially in the SP dataset) and the role of the inertia in explaining current choices.

  6. Application of cumulative prospect theory for the estimation of reliability by Ramjerdi Farideh (Institute of Transport Economics)
  7. Metropolis-Hastings sampling of alternatives for route choice models by Gunnar Flötteröd (EPFL TRANSP-OR)

    Sampling of alternatives has been shown to be an operational technique for the estimation of route choice models. In order to apply it, the sampling protocol generating paths must be such that the sampling probability is known, in order to correct for the sampling bias. Few path generation algorithms proposed in the literature allow to compute the associated probability.

    We present a new path generation method that we are currently investigating. It is based on the Metropolis-Hastings algorithm, so that the sampling probability is actually an input to the method, and not an output. Also, it allows for a great deal of flexibility in terms of the exploration of the set of paths.

  8. Analysis of the impact of travelers' attitudes and perceptions on mode choice in low-density areas by Aurélie Glerum (EPFL)

    Mode choice is influenced by quantitative aspects such as cost and time, but importance should be given to qualitative aspects such as people's attitudes or perceptions towards the different modes as well. These latent characteristics can be integrated into a discrete choice model. In this presentation, we will investigate the effect of attitudes against public transports, environmental concerns or travelers' images of the transportation modes on the travelers' mode choice

  9. Bootstrapping approach for sampling of alternatives in MEV models by Ricardo Hurtubia (Ecole Polytechnique Fédérale de Lausanne)

    Sampling of alternatives is easily implemented in multinomial logit models; however, the case is not the same for more complex models. Ben-Akiva and Guevara (2010) proposed an asymptotically unbiased estimator for MEV models with sampling of alternatives. We investigate to what extent the bootstrapping technique can be used to reduce this bias for small sample sizes on the MEV formulation of a nested logit model.

  10. The Spitsmijden experiments: the impact of rewards on departure time choice by Jasper Knockaert (VU University Amsterdam)

    A series of Spitsmijden pricing experiments are being conducted in the Netherlands. The different experiments focus on the use of positive monetary stimuli (rewards) to change travel behaviour in order to reduce peak demand, both for road and rail commute. Both revealed and stated preference data are collected.

    The presentation will provide an overview of the experiments and discuss the datasets collected and the methodological approach in the analyses of departure time choice using discrete choice theory.

  11. Modeling the behavior of investors by Thomas Robin (TRANSP-OR, EPFL)

    The ability to predict the evolution of financial markets has motivated many researches. This evolution depends on the decisions taken by some financial actors. Few models have been proposed to handle with the disagregate behavior of such actors. We propose to understand and model the behavior of the investors. The predictions of the model could be used to help decision makers and simulate the evolution of markets. Two case studies will be presented, including the data and the adopted discrete choice approaches.

List of participants (15)

Top
  • Atasoy Bilge, EPFL
  • Axhausen Kay, ETH Zürich
  • Ben-Elia Eran, University of West of England
  • Bierlaire Michel, EPFL
  • Chen Jingmin, EPFL
  • Chen Yu, Transport System Planning and Transport Telematics Laboratory, Berlin Institute of Technology
  • Cherchi Elisabetta, University of Cagliari
  • Farideh Ramjerdi, Institute of Transport Economics
  • Flötteröd Gunnar, EPFL TRANSP-OR
  • Glerum Aurélie, EPFL
  • Hurtubia Ricardo, Ecole Polytechnique Fédérale de Lausanne
  • Knockaert Jasper, VU University Amsterdam
  • Robin Thomas, TRANSP-OR, EPFL
  • Viswanathan Prem Kumar, EPFL
  • Wang Yun-Pang, Institute of Transporation Systems (DLR)