Filters








17 Hits in 1.3 sec

Hybrid Repeat/Multi-point Sampling for Highly Volatile Objective Functions [article]

Brett Israelsen, Nisar Ahmed
2016 arXiv   pre-print
A key drawback of the current generation of artificial decision-makers is that they do not adapt well to changes in unexpected situations. This paper addresses the situation in which an AI for aerial dog fighting, with tunable parameters that govern its behavior, will optimize behavior with respect to an objective function that must be evaluated and learned through simulations. Once this objective function has been modeled, the agent can then choose its desired behavior in different situations.
more » ... Bayesian optimization with a Gaussian Process surrogate is used as the method for investigating the objective function. One key benefit is that during optimization the Gaussian Process learns a global estimate of the true objective function, with predicted outcomes and a statistical measure of confidence in areas that haven't been investigated yet. However, standard Bayesian optimization does not perform consistently or provide an accurate Gaussian Process surrogate function for highly volatile objective functions. We treat these problems by introducing a novel sampling technique called Hybrid Repeat/Multi-point Sampling. This technique gives the AI ability to learn optimum behaviors in a highly uncertain environment. More importantly, it not only improves the reliability of the optimization, but also creates a better model of the entire objective surface. With this improved model the agent is equipped to better adapt behaviors.
arXiv:1612.03981v1 fatcat:ja65uc2qajcpzadveq3kcy57jy

Generalized Laguerre Reduction of the Volterra Kernel for Practical Identification of Nonlinear Dynamic Systems [article]

Brett W. Israelsen, Dale A. Smith
2014 arXiv   pre-print
The Volterra series can be used to model a large subset of nonlinear, dynamic systems. A major drawback is the number of coefficients required model such systems. In order to reduce the number of required coefficients, Laguerre polynomials are used to estimate the Volterra kernels. Existing literature proposes algorithms for a fixed number of Volterra kernels, and Laguerre series. This paper presents a novel algorithm for generalized calculation of the finite order Volterra-Laguerre (VL) series
more » ... for a MIMO system. An example addresses the utility of the algorithm in practical application.
arXiv:1410.0741v1 fatcat:plhkl5qcgva2zfs53fnhcpjsby

Towards Adaptive Training of Agent-based Sparring Partners for Fighter Pilots [article]

Brett W. Israelsen, Nisar Ahmed, Kenneth Center, Roderick Green, Winston Bennett Jr
2016 arXiv   pre-print
A key requirement for the current generation of artificial decision-makers is that they should adapt well to changes in unexpected situations. This paper addresses the situation in which an AI for aerial dog fighting, with tunable parameters that govern its behavior, must optimize behavior with respect to an objective function that is evaluated and learned through simulations. Bayesian optimization with a Gaussian Process surrogate is used as the method for investigating the objective function.
more » ... One key benefit is that during optimization, the Gaussian Process learns a global estimate of the true objective function, with predicted outcomes and a statistical measure of confidence in areas that haven't been investigated yet. Having a model of the objective function is important for being able to understand possible outcomes in the decision space; for example this is crucial for training and providing feedback to human pilots. However, standard Bayesian optimization does not perform consistently or provide an accurate Gaussian Process surrogate function for highly volatile objective functions. We treat these problems by introducing a novel sampling technique called Hybrid Repeat/Multi-point Sampling. This technique gives the AI ability to learn optimum behaviors in a highly uncertain environment. More importantly, it not only improves the reliability of the optimization, but also creates a better model of the entire objective surface. With this improved model the agent is equipped to more accurately/efficiently predict performance in unexplored scenarios.
arXiv:1612.04315v1 fatcat:qod7jlzkv5byhdqwk5mzsipk7u

Factorized Machine Self-Confidence for Decision-Making Agents [article]

Brett W Israelsen, Nisar R Ahmed, Eric Frew, Dale Lawrence, Brian Argrow
2019 arXiv   pre-print
Algorithmic assurances from advanced autonomous systems assist human users in understanding, trusting, and using such systems appropriately. Designing these systems with the capacity of assessing their own capabilities is one approach to creating an algorithmic assurance. The idea of 'machine self-confidence' is introduced for autonomous systems. Using a factorization based framework for self-confidence assessment, one component of self-confidence, called 'solver-quality', is discussed in the
more » ... ntext of Markov decision processes for autonomous systems. Markov decision processes underlie much of the theory of reinforcement learning, and are commonly used for planning and decision making under uncertainty in robotics and autonomous systems. A 'solver quality' metric is formally defined in the context of decision making algorithms based on Markov decision processes. A method for assessing solver quality is then derived, drawing inspiration from empirical hardness models. Finally, numerical experiments for an unmanned autonomous vehicle navigation problem under different solver, parameter, and environment conditions indicate that the self-confidence metric exhibits the desired properties. Discussion of results, and avenues for future investigation are included.
arXiv:1810.06519v2 fatcat:abj7xeeusbbodp64pp3xmbizoi

Adaptive Simulation-based Training of AI Decision-makers using Bayesian Optimization [article]

Brett W. Israelsen, Nisar Ahmed, Kenneth Center, Roderick Green, Winston Bennett Jr
2017 arXiv   pre-print
This work studies how an AI-controlled dog-fighting agent with tunable decision-making parameters can learn to optimize performance against an intelligent adversary, as measured by a stochastic objective function evaluated on simulated combat engagements. Gaussian process Bayesian optimization (GPBO) techniques are developed to automatically learn global Gaussian Process (GP) surrogate models, which provide statistical performance predictions in both explored and unexplored areas of the
more » ... r space. This allows a learning engine to sample full-combat simulations at parameter values that are most likely to optimize performance and also provide highly informative data points for improving future predictions. However, standard GPBO methods do not provide a reliable surrogate model for the highly volatile objective functions found in aerial combat, and thus do not reliably identify global maxima. These issues are addressed by novel Repeat Sampling (RS) and Hybrid Repeat/Multi-point Sampling (HRMS) techniques. Simulation studies show that HRMS improves the accuracy of GP surrogate models, allowing AI decision-makers to more accurately predict performance and efficiently tune parameters.
arXiv:1703.09310v2 fatcat:ppmame2e2vhdjnzxzqaxnkdqpu

"I can assure you [...] that it's going to be all right" -- A definition, case for, and survey of algorithmic assurances in human-autonomy trust relationships [article]

Brett W Israelsen
2017 arXiv   pre-print
[13] , and Israelsen et al. [57] ). Representation learning and Feature Selection.  ... 
arXiv:1708.00495v2 fatcat:bx42oqcijfhchna3xhl7nktnqa

"Dave...I can assure you ...that it's going to be all right ..." A Definition, Case for, and Survey of Algorithmic Assurances in Human-Autonomy Trust Relationships

Brett W. Israelsen, Nisar R. Ahmed
2019 ACM Computing Surveys  
Israelsen and Nisar R. Ahmed tasks), and 3) re nement (ability to improve performance on tasks).  ... 
doi:10.1145/3267338 fatcat:7wzpdawvwbdmvavttqfsu2fly4

The relationship between axial resolution and signal-to-noise ratio in optical coherence tomography [article]

Danielle J. Harper, Benjamin J. Vakoc
2021 arXiv   pre-print
[6] Martin Villiger, Ellen Ziyi Zhang, Seemantini K Nadkarni, Wang-Yuhl Oh, Benjamin J Vakoc, and Brett E Bouma.  ...  Lett., 45(7):2107–2110, 2020. [13] Mikkel Jensen, Iván Bravo Gonzalo, Rasmus Dybbro Engelsholm, Michael Maria, Niels Møller Israelsen, Adrian Podoleanu, and Ole Bang.  ... 
arXiv:2112.09226v1 fatcat:vfuyug34d5afzkm7ilqlalxtai

Annual Variation of Spawning Cutthroat Trout in a Small Western USA Stream: A Case Study with Implications for the Conservation of Potamodromous Trout Life History Diversity

Stephen Bennett, Robert Al-Chokhachy, Brett B. Roper, Phaedra Budy
2014 North American Journal of Fisheries Management  
Both resident and fluvial life history forms of Bonneville Cutthroat Trout spawn in Spawn Creek (Bernard and Israelsen 1982; Budy et al. 2007 ).  ...  -We suspected that beaver dams in Spawn Creek (Bernard and Israelsen 1982; Lokteff et al. 2013 ) could influence the distribution of redds as changes in flows related to the construction and failure of  ... 
doi:10.1080/02755947.2014.938139 fatcat:fs4ykumwmne5xeo62ywbe43ple

Riparian vegetation communities change rapidly following passive restoration at a northern Utah stream

Nate Hough-Snee, Brett B. Roper, Joseph M. Wheaton, Phaedra Budy, Ryan L. Lokteff
2013 Ecological Engineering: The Journal of Ecotechnology  
Because Spawn Creek is important cutthroat trout spawning habitat (Bernard and Israelsen, 1982) , passive riparian restoration was initially undertaken to increase vegetation density and abundance to  ... 
doi:10.1016/j.ecoleng.2013.07.042 fatcat:dhoxpi5hg5eohc3z65yhk7e5eq

Do Beaver Dams Impede the Movement of Trout?

Ryan L. Lokteff, Brett B. Roper, Joseph M. Wheaton
2013 Transactions of the American Fisheries Society  
Recently, spawning activity by Bonneville Cutthroat Trout has increased within the 75 m above the pool of S2 (Brett B. Roper, unpublished data).  ...  in the upper portion of Spawn Creek are part of a major beaver dam complex that has been Downloaded by [National Forest Service Library] at 08:49 26 June 2013 present for over 30 years (Bernard and Israelsen  ... 
doi:10.1080/00028487.2013.797497 fatcat:tfk72mw4sfd2pfqv5ul74oau5q

Aligning Conservation Goals and Management Objectives for Bonneville Cutthroat Trout (Oncorhynchus Clarki Utah) in the Logan River, Utah

Harrison Mohn
2016
Brett Roper and Phaedra Budy, for their continuous support and guidance throughout the entire process.  ...  Observed movement patterns indicate that these streams could be particularly important for production of fish that will inhabit the lower watershed (Bernard and Israelsen 1982) .  ... 
doi:10.26076/8c41-5964 fatcat:4ekonhaoqbayjkjjadywsobhe4

Proximal nitrogen reduces the fluorescence quantum yield of nitrogen-vacancy centres in diamond [article]

Marco Capelli, Lukas Lindner, Tingpeng Luo, Jan Jeske, Hiroshi Abe, Shinobu Onoda, Takeshi Ohshima, Brett Johnson, David A. Simpson, Alastair Stacey, Philipp Reineck, Brant C. Gibson (+1 others)
2021 arXiv   pre-print
Proximal nitrogen reduces the fluorescence quantum yield of nitrogen-vacancy centres in diamond Marco Capellia, Lukas Lindnerb, Tingpeng Luob, Jan Jeskeb, Hiroshi Abec, Shinobu Onodac, Takeshi Ohshimac, Brett  ...  Radko IP, Boll M, Israelsen NM, Raatz N, Meijer J, Jelezko F, Andersen UL, Huck A (2016) Determining the internal quantum efficiency of shallow-implanted nitrogen-vacancy defects in bulk diamond  ... 
arXiv:2112.07260v1 fatcat:44ztsbd4jrhbzhuiwk4r4uvhbi

A Study of the Spawning Ecology and Early Life History Survival of Bonneville Cutthroat Trout

Phaedra Budy, Sara Wood, Brett Roper
2012 North American Journal of Fisheries Management  
In addition, Spawn Creek, anecdotally known as a primary area for Bonneville cutthroat trout spawning (Fleener 1951; Bernard and Israelsen 1982) , was the site of a recent restoration project, where a  ... 
doi:10.1080/02755947.2012.675945 fatcat:ixdvpkndbvhklccqjqoyum5n5e

What Does the Public Get? Experimental Use and the Patent Bargain

Katherine J. Strandburg
2003 Social Science Research Network  
Israelsen, Making, Using, and Selling Without Infringing: An Examination of 35 U.S.C.  ...  See Brett Frischmann, Innovation and Institutions: Rethinking the Economics of U.S. Science and Technology Policy, 24 VT. L. REV. 347, 350-53, 386-91 (2000).  ... 
doi:10.2139/ssrn.438023 fatcat:iykoe246evf5dmlglrz47jvlvm
« Previous Showing results 1 — 15 out of 17 results