Detection of fraud, waste, and abuse (FWA) is an important yet challenging problem. In this article, we describe a system to detect suspicious activities in large health‐care data sets. Each healthcare data set is viewed as a heterogeneous network consisting of millions of patients, hundreds of thousands of doctors, tens of thousands of pharmacies, and other entities. Graph‐analysis techniques are developed to find suspicious individuals, suspicious relationships between individuals, unusual changes over time, unusual geospatial dispersion, and anomalous network structure. The visualization interface, known as the network explorer, provides a good overview of data and enables users to filter, select, and zoom into network details on demand. The system has been deployed on multiple sites and data sets, both government and commercial, and identified many overpayments with a potential value of several million dollars per month.
This paper investigates ways to obtain consumer preferences for technology products to help designers identify the key attributes that contribute to a product's market success. A case study of residential photovoltaic panels is performed in the context of the California, USA, market within the 2007–2011 time span. First, interviews are conducted with solar panel installers to gain a better understanding of the solar industry. Second, a revealed preference method is implemented using actual market data and technical specifications to extract preferences. The approach is explored with three machine learning methods: Artificial neural networks (ANN), Random Forest decision trees, and Gradient Boosted regression. Finally, a stated preference self-explicated survey is conducted, and the results using the two methods compared. Three common critical attributes are identified from a pool of 34 technical attributes: power warranty, panel efficiency, and time on market. From the survey, additional nontechnical attributes are identified: panel manufacturer's reputation, name recognition, and aesthetics. The work shows that a combination of revealed and stated preference methods may be valuable for identifying both technical and nontechnical attributes to guide design priorities.
Large-scale engineering systems require design teams to balance complex sets of considerations using a wide range of design and decision-making skills. Formal, computational approaches for optimizing complex systems offer strategies for arriving at optimal solutions in situations where system integration and design optimization are well-formulated. However, observation of design practice suggests engineers may be poorly prepared for this type of design. Four graduate student teams completed a distributed, complex system design task. Analysis of the teams' design histories suggests three categories of suboptimal approaches: global rather than local searches, optimizing individual design parameters separately, and sequential rather than concurrent optimization strategies. Teams focused strongly on individual subsystems rather than system-level optimization, and did not use the provided system gradient indicator to understand how changes in individual subsystems impacted the overall system. This suggests the need for curriculum to teach engineering students how to appropriately integrate systems as a whole.
12 July 2010 in Artificial Intelligence for Engineering Design, Analysis and Manufacturing
User-centered approaches to design can guide teams toward an understanding of users and aid teams in better posing design problems. This paper investigates the role of user-centered design approaches in design process and outcome within the context of design team projects. The value of interaction with users is examined at several stages throughout the design process. The influence of user-centered design on the performance of design teams is also explored. Results suggest that the quantity of interactions with users and time spent interacting with users alone is not linked with better design outcome, but that iterative evaluation of concepts by users may be of particular value to design prototypes. Suggestions are made based on the reflections from the authors after conducting this study.