Daniel Dolk

Daniel Dolk

Auto-generated profile, based on publicly available information

Daniel Dolk

Naval Postgraduate School

Research Focus & Metrics

Academic Career Length

36Years (1991-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
13 December 2011 in Information Systems and e-Business Management
Verified
Journal Article
Closed access
PDF file
DKDon Kridel
Don Kridel
DDDaniel Dolk
Daniel Dolk

1 January 2023
Verified
Conference Paper
Open access
PDF file
JDJacob Dineen
Jacob Dineen
DKDon Kridel
Don Kridel
DDDaniel Dolk
Daniel Dolk
DCDavid Castillo
David Castillo
A high-velocity paradigm shift towards Explainable Artificial Intelligence (XAI) has emerged in recent years. Highly complex Machine Learning (ML) models have flourished in many tasks of intelligence, and the questions have started to shift away from traditional metrics of validity towards something deeper: What is this model telling me about my data, and how is it arriving at these conclusions? Previous work has uncovered predictive models generating explanations contrasting domain experts, or excessively exploiting bias in data that renders a model useless in highly-regulated settings. These inconsistencies between XAI and modeling techniques can have the undesirable effect of casting doubt upon the efficacy of these explainability approaches. To address these problems, we propose a systematic, perturbation-based analysis against a popular, model-agnostic method in XAI, SHapley Additive exPlanations (Shap). We devise algorithms to generate relative feature importance in settings of dynamic inference amongst a suite of popular machine learning and deep learning methods, and metrics that allow us to quantify how well explanations generated under the static case hold. We propose a taxonomy for feature importance methodology, measure alignment, and observe quantifiable similarity amongst explanation models across several datasets.

20 October 2005 in Information Systems and e-Business Management
Verified
Journal Article
Closed access
PDF file
ACAlok R. Chaturvedi
Alok R. Chaturvedi
SMShailendra Raj Mehta
Shailendra Raj Mehta
DDDaniel Dolk
Daniel Dolk
MGMukul Gupta
Mukul Gupta

22 March 2012 in Information Systems and e-Business Management
Verified
Journal Article
Closed access
PDF file
HDHaluk Demirkan
Haluk Demirkan
DDDaniel Dolk
Daniel Dolk

1 April 1993 in Interfaces
Verified
Journal Article
Closed access
DKDonald J. Kridel
Donald J. Kridel
DDDaniel R. Dolk
Daniel R. Dolk
Over the last 20 years, legal and regulatory decisions, culminating in the divestiture of the Bell Operating Companies (BOCs) by American Telephone and Telegraph (AT&T) in 1984, have led to an upsurge in the importance of applied econometrics in the telecommunications industry. Increased attention to demand analysis, or empirical economics, has fostered a commensurate increase in modeling sophistication—both for the analyst and for the underlying modeling component of decision support systems. Whereas (aggregate) pooled regression models were more appropriate for the rate-case environment, (disaggregate) optimization-based models have proved more useful for the marketing orientation. Stand-alone analytical systems are no longer sufficient. The critical transition in organizational perspective reflects a need for integrated modeling environments that offer multiple modeling paradigms, linkage among models, consistent access to data resources, and intelligent support.

18 December 1991 in Journal of Management Information Systems
Verified
Journal Article
Closed access
DCDavid G. Castillo
David G. Castillo
DDDaniel R. Dolk
Daniel R. Dolk
DKDonald J. Kridel
Donald J. Kridel
This paper describes an active, or symbiotic, decision support system developed for the National Aeronautics and Space Administration called the Generic Operations Simulation Technique (GOST). GOST combines parametric modeling techniques (regression) with discrete-event simulation to cost and plan future space programs. GOST differs from previous process modeling approaches by providing an “artificially intelligent modeling expert” and an “artificially intelligent domain expert” for assisting the user in developing and analyzing process models. An object oriented knowledge representation system provides the foundation for the GOST environment. A taxonomy of Computer Directed Process Managers, implemented within the representation system, controls the appropriate processes responsible for delivering active modeling support.