Stochastic similarity for validating human control strategy models

Nechyba, M.C.; Xu, Yangsheng

doi:10.1109/robot.1997.620051

Cited by 23 publications

(28 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As is shown in [18], this metric is quite well-behaved as it exhibits the following useful properties:…”

Section: A Methodologymentioning

confidence: 94%

“…We have adopted a stochastic method of model validation introduced in dynamic human control strategy modeling [18]. This method of model validation utilizes hidden markov model optimization via the Baum-Welch expectation-maximization (EM) algorithm.…”

Section: A Methodologymentioning

confidence: 99%

“…We use the obvious choice of 4 hidden states, but also tried many experiments with 8 states as well (the most used in [18]) to see if there was notable variance -there was not.…”

Section: A Methodologymentioning

confidence: 99%

See 2 more Smart Citations

An Empirical Activity Model for WLAN Users

Phillips

Singh

2008

IEEE INFOCOM 2008 - The 27th Conference on Computer Communications

View full text Add to dashboard Cite

Abstract-Understanding user behavior in wireless environments is useful for a variety of reasons ranging from the design of better sleep algorithms for components of mobile devices to appropriately provisioning the wireless network itself to better serve the user. Our work goes in a different direction from prior work on WLAN modeling and attempts to undersand the protocol independent behavior of users by developing packet-level models for user activity using diverse training data. Additionally we validate the derived model using a stochastic similarity metric adapted from human control strategy modeling and present a novel way to compare traces using this metric.

show abstract

“…As is shown in [18], this metric is quite well-behaved as it exhibits the following useful properties:…”

Section: A Methodologymentioning

confidence: 94%

Section: A Methodologymentioning

confidence: 99%

See 1 more Smart Citation

An Empirical Activity Model for WLAN Users

Phillips

Singh

2008

IEEE INFOCOM 2008 - The 27th Conference on Computer Communications

View full text Add to dashboard Cite

show abstract

“…Step m: 36-m situations m=m+1 chose the pseudo-distance defined by Nechyba & Xu for the HMM (Nechyba & Xu, 1998). This distance has the advantage of evaluating the proximity between models with different dimensions by computing the model likelihood regarding the learning sequence associated to each one.…”

Section: Group Closest Situationsmentioning

confidence: 99%

Weight Semi Hidden Markov Model and Driving Situation Classification for Driver Behavior Diagnostic

Dapzol¹

2007

Proceedings of the 4th International Driving Symposium on Human Factors in Driver Assessment, Training, and Vehicle

View full text Add to dashboard Cite

Summary:In this study, we propose to use statistical modelling to analyze, model, and categorize driving activity. To achieve this objective, we develop a new statistical model by adding a weight feature to the classic Semi Hidden Markov Model (SHMM) framework. Then, to assess its capacity, we conduct an experiment that allows us to record 718 driving sequences categorized in 36 situations. We then used our modelling to identify the driver's aim and the driving situation he's in. Furthermore, we adapted the ascendant hierarchic classification technique to this modelling. It allows us to understand which situations are close and to define partitions of whole driving situations. Finally, on these sequences, our modelling choice allows us to predict the driver's situation with, on average, an 85% success rate. These results show the HMM effectiveness to manage temporal and multidimensional data by modelling predicting drivers' behavior.

show abstract

“…In previous work, a stochastic similarity measure, which compares model-generated control trajectories to the original human training data, has been proposed for validating HCS models [9]. While this similarity measure can ensure that a given HCS model adequately captures the driving characteristics of the human operator, it does not measure a particular model's skill or performance.…”

Section: Introductionmentioning

confidence: 99%

Performance evaluation and optimization of human control strategy

Song

Nechyba

et al. 2002

Robotics and Autonomous Systems

View full text Add to dashboard Cite

Modeling human control strategy (HCS) is becoming an increasingly popular paradigm in a number of different research areas, ranging from robotics and intelligent vehicle highway systems to expert training and virtual reality computer games. Usually, HCS models are derived empirically, rather than analytically, from real-time human input-output data. While these empirical models offer an effective means of transferring intelligent behaviors from humans to robots and other machines, there is a great need to develop adequate performance criteria for these models. It is our goal in this paper to develop several such criteria for the task of human driving. We first collect driving data from different individuals through a real-time graphic driving simulator that we have developed, and identify each individual's control strategy model through the flexible cascade neural network learning architecture. We then define performance measures for evaluating two aspects of the resultant HCS models. The first is based on event analysis, while the second is based on inherent analysis. Using the proposed performance criteria, we demonstrate the procedure for evaluating the relative skill of different HCS models. Finally, we propose an iterative algorithm for optimizing an initially stable HCS model with respect to independent, user-specified performance criteria, by applying the simultaneously perturbed stochastic approximation (SPSA) algorithm. The methods proposed herein offer a means for modeling and transferring HCS in response to real-time inputs, and improving the intelligent behaviors of artificial machines.

show abstract

Stochastic similarity for validating human control strategy models

Abstract: Modeling dynamic human control strategy (HCS)

Cited by 23 publications

References 4 publications

An Empirical Activity Model for WLAN Users

An Empirical Activity Model for WLAN Users

Weight Semi Hidden Markov Model and Driving Situation Classification for Driver Behavior Diagnostic

Performance evaluation and optimization of human control strategy

Contact Info

Product

Resources

About