Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Real-Time Pricing Optimization for Ride-Hailing Quality of Service
Authors: Enpeng Yuan, Pascal Van Hentenryck
IJCAI 2021 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Simulation experiments indicate that the pricing optimization model achieves short waiting times without sacrificing revenues or geographical fairness. The performance of AP-RTRS is evaluated using Yellow Taxi trip data in Manhattan, New York City [NYC, 2019a]. The experiments compare three MPC models: RELOCATION, SURGE, and SURGE+POSPTPONE. Table 2 reports the dropout rate, the number of riders served, and the waiting times. |
| Researcher Affiliation | Academia | Enpeng Yuan , Pascal Van Hentenryck Georgia Institute of Technology, Atlanta, USA EMAIL |
| Pseudocode | No | The paper presents a mathematical formulation in Figure 2 ('The MPC Optimization with Pricing and Relocation'), but it is an optimization model (MILP) and not pseudocode or a clearly labeled algorithm block. |
| Open Source Code | No | The paper does not include any explicit statement about releasing source code for the described methodology, nor does it provide links to a code repository. |
| Open Datasets | Yes | The performance of AP-RTRS is evaluated using Yellow Taxi trip data in Manhattan, New York City [NYC, 2019a]. [NYC, 2019a] NYC. Nyc taxi & limousine commission trip record data. https://www1.nyc.gov/site/tlc/passengers/ taxi-fare.page, 2019. Accessed: 2020-10-01. |
| Dataset Splits | No | The paper mentions 'The test data is generated based on Yellow Taxi trip data', but it does not specify explicit training, validation, or test dataset splits (e.g., percentages, sample counts, or references to predefined splits) needed for reproducibility. |
| Hardware Specification | Yes | The model is solved (optimally or near optimally) by Gurobi 9.0 in 30 seconds with 32 CPU cores [Gurobi Optimization, 2020]. |
| Software Dependencies | Yes | The model is solved (optimally or near optimally) by Gurobi 9.0 in 30 seconds with 32 CPU cores [Gurobi Optimization, 2020]. |
| Experiment Setup | Yes | Each request in the simulation is given a maximum scheduling time of 5 minutes and a maximum waiting time of 15 minutes. The pricing model is run every 5 minutes and has a time horizon of T = 4 epochs. Demand predictions for each O-D pair in each epoch is generated by adding white noise to the true demand. The white noise is normally distributed with zero mean and a standard deviation equal to 2.5% of the true demand. Ride-share ratio is set to be Wij = 1.4 for all i, j Z. Service weight and relocation penalty functions are as follows: qp(t, τ, ρ) = 0.5t0.75τ t0.67ρ τ, and qr(t) = 0.001 0.5t. The model evaluation considers three elasticity levels: 0.5, 1.0, and 2.0 for SURGE. |