Scianta Intelligence White Papers

How can we discover anomalous behaviors in very large systems with high degrees of noise and time-varying attributes? This white paper is concerned with discovering provider fraud and abuse in managed healthcare and is based on find and exploiting normal peer behaviors in large populations of providers with similar properties.

Covering issues and techniques in cluster analysis and self-organizing maps as well as statistical learning theory, this paper is intended for business analysts, knowledge engineers, and technically inclined managers who need to understand ways of looking deeply into large volume of high dimensional, volatile data and discovering normal and abnormal behavior patterns.

The fusion of object oriented design, machine intelligence, and a distributed networking superstructure provides a powerful and robust paradigm for the delivery of fault-tolerant, self-tuning, and adaptable applications. Using an set of examples from retailing, this white paper explains how intelligent objects can enhance an organization’s competitive edge, improve customer satisfaction, reduce inventory, and manage a distributed supply chain.

Intelligence and adaptability provides a strong framework for applications that can respond to external pressures, learn the behaviors of customers thus reducing customer support, re-programming, maintenance, as well as extending the useful life of existing systems.

What are the different kinds of predictive modeling, how are they used, and what benefits can we gain from the various families of predictive models? This white paper explores the nature and benefits of predicting the future through classification and regression while providing some insights into the properties and characteristics we should look for in a predictive model.

This is an introductory paper for analysts and managers interested in understanding the scope as well as the limitations of predictive models as well as for eBusiness managers interested in exploring ways to predict sales, web site traffic, inventory volumes, and similar key performance indicators.

We live in a highly connected world. These connections evolve as emergent behaviors from the existence of smaller, oftentimes loosely connected, networks. How can we find, understand, quantify, and exploit these linkages to better understand the behavior of the underlying systems? Answering this question is a crucial element in customer segmentation, pricing models, campaign development, new line of business development as well as our continued fight against terrorism.

This white paper, part one in a series on Link Analysis, explores the nature of networks, establishes a basic vocabulary and working knowledge of graphs, and illustrates linkage analysis through several real-world examples