Trajectory Prediction from a Mass of Sparse and Missing External Sensor Data

Cruz, Lívia Almada; Zeitouni, Karine; Macêdo, José Antônio Fernandes de

doi:10.1109/mdm.2019.00-43

Cited by 3 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is also worth noting that the uninformative RU-IRL method achieved good performance despite being much less flexible, with only two estimated parameters (𝛽 1 and 𝛽 2 ), while the number of estimated parameters of the Markov method corresponds to the number of transition probabilities (272 × 271 possible transitions). This is also in accordance with results obtained by [4], in which a recurrent neural network with both location and timestamp inputs achieved only a slightly better performance than a first-order Markov model in the next location prediction task.…”

Section: Case Studysupporting

confidence: 92%

“…In [4], the authors propose a recurrent neural network model to predict the next location from moving object trajectories captured by external sensors (e.g., traffic surveillance cameras) placed on the roadside. They also cope with the incompleteness and sparsity problems that are inherent to trajectories captured by sensors, and propose a scheme to integrate the solutions to such problems into the prediction model.…”

Section: Related Workmentioning

confidence: 99%

“…In this paper, we consider the problem of modeling trajectories of vehicles in a road network which are observed by external sensors located on sparse fixed points on the roadside. In contrast to the majority of previous work on trajectory modeling, in which trajectories are made up of GPS traces, trajectories from external sensors are more sparse and noisy, which makes the problem of modeling trajectories more challenging [3,4]. While GPS traces have a high sample rate, which allows us to model trajectories as varying almost continuously over the space, external sensors are placed in fixed locations and observations from the same vehicle may be very far apart.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Trajectory Modeling via Random Utility Inverse Reinforcement Learning

Pitombeira‐Neto¹,

Santos²,

Silva³

et al. 2021

Preprint

View full text Add to dashboard Cite

We consider the problem of modeling trajectories of drivers in a road network from the perspective of inverse reinforcement learning. As rational agents, drivers are trying to maximize some reward function unknown to an external observer as they make up their trajectories. We apply the concept of random utility from microeconomic theory to model the unknown reward function as a function of observable features plus an error term which represents features known only to the driver. We develop a parameterized generative model for the trajectories based on a random utility Markov decision process formulation of drivers decisions. We show that maximum entropy inverse reinforcement learning is a particular case of our proposed formulation when we assume a Gumbel density function for the unobserved reward error terms. We illustrate Bayesian inference on model parameters through a case study with real trajectory data from a large city obtained from sensors placed on sparsely distributed points on the street network.

show abstract

Section: Case Studysupporting

confidence: 92%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Trajectory Modeling via Random Utility Inverse Reinforcement Learning

Pitombeira‐Neto¹,

Santos²,

Silva³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

Trajectory modeling via random utility inverse reinforcement learning

Pitombeira-Neto,

Santos,

Coelho da Silva

et al. 2024

Information Sciences

View full text Add to dashboard Cite

Modeling Trajectories Obtained from External Sensors for Location Prediction via NLP Approaches

Cruz

Silva

Magalhães

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

Representation learning seeks to extract useful and low-dimensional attributes from complex and high-dimensional data. Natural language processing (NLP) was used to investigate the representation learning models to extract words’ feature vectors using their sequential order in the text via word embeddings and language models that maintain their semantic meaning. Inspired by NLP, in this paper, we tackle the representation learning problem for trajectories, using NLP methods to encode external sensors positioned in the road network and generate the features’ space to predict the next vehicle movement. We evaluate the vector representations of on-road sensors and trajectories using extrinsic and intrinsic strategies. Our results have shown the potential of natural language models to describe the space of features on trajectory applications as the next location prediction.

show abstract

Trajectory Prediction from a Mass of Sparse and Missing External Sensor Data

Cited by 3 publications

References 12 publications

Trajectory Modeling via Random Utility Inverse Reinforcement Learning

Trajectory Modeling via Random Utility Inverse Reinforcement Learning

Trajectory modeling via random utility inverse reinforcement learning

Modeling Trajectories Obtained from External Sensors for Location Prediction via NLP Approaches

Contact Info

Product

Resources

About