Neural Approximate Dynamic Programming for On-Demand Ride-Pooling
Authors: Sanket Shah, Meghna Lowalekar, Pradeep Varakantham507-515
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In the experiments, we compare our approach to two leading approaches for solving the RMP on a real-world dataset(NYYellow Taxi 2016). Compared to a baseline approach proposed by (Alonso-Mora et al. 2017), we show that our approach serves up to 16% more seen requests across different parameter settings. |
| Researcher Affiliation | Academia | Sanket Shah, Meghna Lowalekar, Pradeep Varakantham School of Information Systems, Singapore Management University {sankets, pradeepv}@smu.edu.sg, meghnal.2015@phdcs.smu.edu.sg |
| Pseudocode | Yes | Algorithm 1: Neur ADP (N, T) |
| Open Source Code | Yes | The setup, code and supplementary file are available at https://github.com/sanketkshah/NeurADP-for-Ride-Pooling. |
| Open Datasets | Yes | The experiments are conducted by taking the demand distribution from the publicly available New York Yellow Taxi Dataset (NYYellow Taxi 2016). NYYellow Taxi. 2016. New york yellow taxi dataset. http://www.nyc.gov/html/tlc/html/about/trip_record_data.shtml. |
| Dataset Splits | Yes | Neur ADP is trained using the data for 8 weekdays (23 March – 1 April 2016) and it is validated on 22 March 2016. |
| Hardware Specification | Yes | All experiments are run on 24 core 2.4GHz Intel Xeon E5-2650 processor and 256GB RAM. |
| Software Dependencies | Yes | The algorithms are implemented in python and optimisation models are solved using CPLEX 12.8. |
| Experiment Setup | Yes | The value of maximum allowable detour delay λ is taken as 2 τ. The decision epoch duration Δ is taken as 60 seconds. During training, we perform exploration by adding Gaussian noise to the predicted Vi values (Plappert et al. 2017). The specifics of the neural network architecture and training can be found in the supplementary file. |