Search

Published Thesis, Environmental pollution caused by mines within the district of Lejweleputswa in Free State is a major contributor to health issues and the inability to grow crops within the mining communities. Mining industries continue to develop environmental management systems/plans to mitigate the impact their operations has on the society. Even with these plans, there are still issues of environmental pollution affecting the society. Though there are Information Communication and Technology (ICT) based pollution monitoring solutions, their use is dismal due to lack of appreciation or understanding of how they disseminate information. Furthermore, non-adopting community members are being regarded as inherently conservative or irrational, but these community members argue that the recommendations and technologies brought to them are not always appropriate to their circumstances. There was concern that local people’s knowledge of their environment, farming systems, and their social as well as economic situation had been ignored and underestimated when ICTs solutions are being implemented (Warburton & Martin, 1999). Another challenge is that there is no station to monitor pollution for small communities such as Nyakallong in the district. This result in mining communities depending on their own local knowledge to observe and monitor mining related environmental pollution. However, this local knowledge has never been tested scientifically or analysed to recognize its usability or effectiveness. Mining companies tend to ignore this knowledge from the communities as it is treated like common information with no much scientific value. As a step towards verifying or validating this local knowledge, fuzzy cognitive maps were used to model, analyse and represent this linguistic local knowledge. Although this local knowledge assists in mitigating environmental pollution, incorporating it with scientific knowledge will improve its relevance, trustworthiness and acceptability by majority of

Publishe Thesis, Information and Communication Technology (ICT) has given rise to new technologies and solutions that were not possible a few years ago. One of these new technologies is electronic voting, also known as e-voting, which is the use of computerised equipment to cast a vote. One of the subsets of e-voting is mobile voting (m-voting). M-voting is the use of mobile phones to cast a vote outside the restricted electoral boundaries. Mobile phones are pervasive; they offer connection anywhere, at any time. However, utilising a fast-growing medium such as the mobile phone to cast a vote, poses various new security threats and challenges. Mobile phones utilise equivalent software design used by personal computers which makes them vulnerable or exposed to parallel security challenges like viruses, Trojans and worms. In the past, security solutions for mobile phones encountered several restrictions in practice. Several methods were used; however, these methods were developed to allow lightweight intrusion detection software to operate directly on the mobile phone. Nevertheless, such security solutions are bound to fail securing a device from intrusions as they are constrained by the restricted memory, storage, computational resources, and battery power of mobile phones. This study compared and evaluated two intrusion detection systems (IDSs), namely Snort and Suricata, in order to propose a cloud-based intrusion detection and prevention system (CIDPS) for m-voting in South Africa. It employed simulation as the primary research strategy to evaluate the IDSs. A quantitative research method was used to collect and analyse data. The researcher established that as much as Snort has been the preferred intrusion detection and prevention system (IDPS) in the past, Suricata presented more effective and accurate results close to what the researcher anticipated. The results also revealed that, though Suricata was proven effective enough to protect m-voting while saving the computational res

Information and Communication Technology (ICT) has given rise to new technologies and solutions that were not possible a few years ago. One of these new technologies is electronic voting, also known as e-voting, which is the use of computerised equipment to cast a vote. One of the subsets of e-voting is mobile voting (m-voting). M-voting is the use of mobile phones to cast a vote outside the restricted electoral boundaries. Mobile phones are pervasive; they offer connection anywhere, at any time. However, utilising a fast-growing medium such as the mobile phone to cast a vote, poses various new security threats and challenges. Mobile phones utilise equivalent software design used by personal computers which makes them vulnerable or exposed to parallel security challenges like viruses, Trojans and worms. In the past, security solutions for mobile phones encountered several restrictions in practice. Several methods were used; however, these methods were developed to allow lightweight intrusion detection software to operate directly on the mobile phone. Nevertheless, such security solutions are bound to fail securing a device from intrusions as they are constrained by the restricted memory, storage, computational resources, and battery power of mobile phones. This study compared and evaluated two intrusion detection systems (IDSs), namely Snort and Suricata, in order to propose a cloud-based intrusion detection and prevention system (CIDPS) for m-voting in South Africa. It employed simulation as the primary research strategy to evaluate the IDSs. A quantitative research method was used to collect and analyse data. The researcher established that as much as Snort has been the preferred intrusion detection and prevention system (IDPS) in the past, Suricata presented more effective and accurate results close to what the researcher anticipated. The results also revealed that, though Suricata was proven effective enough to protect m-voting while saving the computational res

Environmental pollution caused by mines within the district of Lejweleputswa in Free State is a major contributor to health issues and the inability to grow crops within the mining communities. Mining industries continue to develop environmental management systems/plans to mitigate the impact their operations has on the society. Even with these plans, there are still issues of environmental pollution affecting the society. Though there are Information Communication and Technology (ICT) based pollution monitoring solutions, their use is dismal due to lack of appreciation or understanding of how they disseminate information. Furthermore, non-adopting community members are being regarded as inherently conservative or irrational, but these community members argue that the recommendations and technologies brought to them are not always appropriate to their circumstances. There was concern that local people’s knowledge of their environment, farming systems, and their social as well as economic situation had been ignored and underestimated when ICTs solutions are being implemented (Warburton & Martin, 1999). Another challenge is that there is no station to monitor pollution for small communities such as Nyakallong in the district. This result in mining communities depending on their own local knowledge to observe and monitor mining related environmental pollution. However, this local knowledge has never been tested scientifically or analysed to recognize its usability or effectiveness. Mining companies tend to ignore this knowledge from the communities as it is treated like common information with no much scientific value. As a step towards verifying or validating this local knowledge, fuzzy cognitive maps were used to model, analyse and represent this linguistic local knowledge. Although this local knowledge assists in mitigating environmental pollution, incorporating it with scientific knowledge will improve its relevance, trustworthiness and acceptability by majority of

Although BNs have been used successfully for many medical diagnosis problems, there have been few applications to epidemiological data where data mining methods play a significant role. In this paper, we look at the application of BNs to epidemiological data, specifically assessment of risk for coronary heart disease (CHD). We build the BNs: (1) by knowledge engineering BNs from two epidemiological models of CHD in the literature; (2) by applying a causal BN learner. We evaluate these BNs using cross-validation. We compared performance in predicting CHD events over 10 years, measuring area under the ROC curve and Bayesian information reward. The knowledge engineered BNs performed as well as logistic regression, while being easier to interpret. These BNs will serve as the baseline in future efforts to extend BN technology to better handle epidemiological data, specifically to model CHD.

The aim of this study has been to find out how the community in the remote areas of South Africa access government information. The study is based on four villages, two in the Northern Cape Province (Heuningvlei and Galeshewe) and two in Limpopo Province (The Oaks and Finala). A quantitative design was used. Open and closed-ended questionnaires were used to collect data from the community. Out of 200 questionnaires distributed only 144 responses were received. Responses in the questionnaires were tabulated, coded and processed using the Statistical Package for Social Sciences (SPSS) programme. Based on the analysis in Table 4.13, the study showed that 44 percent of the sampled community have access to information through Television which they classified as technology. Shortage of telecentres, distance to telecentres, age, lack of education, monthly income, infrastructure and transport cost were some of the important factors contributing to a lack of access to information. Some of the findings are that most of the telecentres are located far from towns, more than 7 kilometres from the village. Ideally, telecentres should be located near the community, within a short walking distance. It is recommended that provision/establishment for easy access to information and communication services by the Universal Service and Access Agency of South Africa (USAASA) in the remote areas in the Northern Cape and Limpopo province is strongly considered for implementation. This will overcome the gap that exists between the urban and semi-urban communities regarding access to human rights information such as rights to life, equality, freedom of speech, assembly and access to information. It is also assumed that the implementation of telecentres and Internet Cafes, in order to facilitate the adoption of e-government information by people residing in remote areas (semi-urban areas) such as Heuningvlei, Galeshewe, The Oaks and Finala will contribute to better access to human rights. The

Thesis ( M. Tech. (School of Information Technology )) - Central University of Technology, Free State, 2014, The aim of this study has been to find out how the community in the remote areas of South Africa access government information. The study is based on four villages, two in the Northern Cape Province (Heuningvlei and Galeshewe) and two in Limpopo Province (The Oaks and Finala). A quantitative design was used. Open and closed-ended questionnaires were used to collect data from the community. Out of 200 questionnaires distributed only 144 responses were received. Responses in the questionnaires were tabulated, coded and processed using the Statistical Package for Social Sciences (SPSS) programme. Based on the analysis in Table 4.13, the study showed that 44 percent of the sampled community have access to information through Television which they classified as technology. Shortage of telecentres, distance to telecentres, age, lack of education, monthly income, infrastructure and transport cost were some of the important factors contributing to a lack of access to information. Some of the findings are that most of the telecentres are located far from towns, more than 7 kilometres from the village. Ideally, telecentres should be located near the community, within a short walking distance. It is recommended that provision/establishment for easy access to information and communication services by the Universal Service and Access Agency of South Africa (USAASA) in the remote areas in the Northern Cape and Limpopo province is strongly considered for implementation. This will overcome the gap that exists between the urban and semi-urban communities regarding access to human rights information such as rights to life, equality, freedom of speech, assembly and access to information. It is also assumed that the implementation of telecentres and Internet Cafes, in order to facilitate the adoption of e-government information by people residing in remote areas (semi-urban areas) such as Heuningvlei, Galeshewe, The Oaks and Finala will contribute to better access to human rights. The

The aim of this study was to create a dynamic software system that captures all information related to a student and delivers it to the educators, principal, higher authorities and parents. In order to achieve this aim, an investigation was launched as to the development of a cost-effective school management system for disadvantaged schools in the Free State Province using the Software as a Service (SaaS) delivery model. Although a variety of other school management systems exist in the market, they are often expensive and difficult to maintain. Details such as previous academic performances, disciplinary actions taken against a student in the current school, ailments the student suffers from and parental details are some of the information that will help an educator to better understand a student. The software that is currently in use fails to deliver this information. Designing the software as a multitenant system, helps accommodate different schools under the same database, while the shared database, shared schema reduces back-end costs. Database design was carried out in such a way that tenant data is logically isolated and that data integrity is maintained throughout. What makes the software explained in this study cost effective is the method of delivery that was employed, which is SaaS. Here, software is not purchased, there is no upfront capital and the yearly license fee is eliminated, as schools need only pay a monthly rental fee for the services they use. Since all services are provided through the Internet, there is no need for system space; the only requirement is a high-speed Internet network.

Thesis (M. Tech. (Information Technology)) -- Central University of technology, Free State, 2013, The aim of this study was to create a dynamic software system that captures all information related to a student and delivers it to the educators, principal, higher authorities and parents. In order to achieve this aim, an investigation was launched as to the development of a cost-effective school management system for disadvantaged schools in the Free State Province using the Software as a Service (SaaS) delivery model. Although a variety of other school management systems exist in the market, they are often expensive and difficult to maintain. Details such as previous academic performances, disciplinary actions taken against a student in the current school, ailments the student suffers from and parental details are some of the information that will help an educator to better understand a student. The software that is currently in use fails to deliver this information. Designing the software as a multitenant system, helps accommodate different schools under the same database, while the shared database, shared schema reduces back-end costs. Database design was carried out in such a way that tenant data is logically isolated and that data integrity is maintained throughout. What makes the software explained in this study cost effective is the method of delivery that was employed, which is SaaS. Here, software is not purchased, there is no upfront capital and the yearly license fee is eliminated, as schools need only pay a monthly rental fee for the services they use. Since all services are provided through the Internet, there is no need for system space; the only requirement is a high-speed Internet network.

Thesis (M. Tech. (Information Technology)) -- Central University of technology, Free State, 2013, Markets for manufactured products are characterized by a fragmentation of the market (with regards to size and time), and by shorter product cycles. This is due to the occurrence of mass customization and globalization. In mass customization, the same basic products are manufactured for a broad market, but then consumers are given the liberty to choose the “finishing touches” that go with the product. The areas that manufacturers now compete for are higher quality products, low cost and timely response to market changes. Appropriate business strategies and manufacturing technologies must thus be used to implement these strategic dimensions. The paradigm of Reconfigurable Manufacturing System (RMS) has been introduced to respond to this new market oriented manufacturing environment. The design of RMS allows ease of reconfiguration as it has a modular structure in terms of software and hardware. This allows ease of reconfiguration as a strategy to adapt to changing market demands. Modularity will allow the ability to integrate/remove software/hardware modules without affecting the rest of the system. RMS can therefore be quickly reconfigured according to the production requirements of new models, it can be quickly adjusted to exact capacity requirements as the market grows and products change, and it is able to integrate new technology. In this research project, real-time, open controller is designed and developed for Reconfigurable Manufacturing Tools (RMTs). RMTs are the basic building blocks for RMS. Real time and openness of the controllers for RMT would allow firstly, for the modular design of RMTs (so that RMTs can be adapted easily for changing product demands) and secondly, prompt control of RMT for diagnosability.

Three progresses were made in the field through mass estimation: (1) the first adaptive version of mass estimation using a new nearest neighbor procedure which runs significantly faster than existing nearest neighbor procedures and needs no indexing schemes, (2) the first mass-based Bayesian classifier which estimates the likelihood directly in multi-dimensional space; unlike existing Bayesian classifiers which estimate simplified surrogates of likelihood (e.g., one-dimensional likelihood), and (3) the first mass-based similarity measure which can be an effective alternative to distance-based similarity measure.

Over the past two decades, machine learning has led to substantial changes in Data Fusion Systems globally. One of the most important application areas for data fusion is situation awareness to support command and control. Situation awareness is perception of elements in the environment, comprehension of the current situation, and projection of future status before decision making. Traditional fusion systems focus on lower levels of the JDL hierarchy, leaving higher-level fusion and situation awareness largely to unaided human judgment. This becomes untenable in today's increasingly data-rich environments, characterized by information and cognitive overload. Higher-level fusion to support situation awareness requires semantically rich representations amenable to automated processing. Ontologies are an essential tool for representing domain semantics and expressing information about entities and relationships in the domain. Probabilistic ontologies augment standard ontologies with support for uncertainty management, which is essential for higher-level fusion to support situation awareness. PROGNOS is a prototype Predictive Situation Awareness (PSAW) System for the maritime domain. The core logic for the PROGNOS probabilistic ontologies is Multi-Entity Bayesian Networks (MEBN), which combine First-Order Logic with Bayesian Networks for representing and reasoning about uncertainty in complex, knowledge-rich domains. MEBN goes beyond standard Bayesian networks to enable reasoning about an unknown number of entities interacting with each other in various types of relationships, a key requirement for PSAW. The existing probabilistic ontology for PROGNOS was constructed manually by a domain expert. However, manual MEBN modeling is labor-intensive and not agile. We have developed a learning algorithm for MEBN-based probabilistic ontologies. This paper presents a bridge between MEBN and the Relational Model, and a parameter and structure learning algorithm for MEBN., Presented at the 18th International Command and Control Research and Technology Symposium (ICCRTS) C2 in Underdeveloped, Degraded and Denied Operational Environments held in Alexandria, VA, on 19-21 June 2013. U.S. Government or Federal Rights License.

Markets for manufactured products are characterized by a fragmentation of the market (with regards to size and time), and by shorter product cycles. This is due to the occurrence of mass customization and globalization. In mass customization, the same basic products are manufactured for a broad market, but then consumers are given the liberty to choose the “finishing touches” that go with the product. The areas that manufacturers now compete for are higher quality products, low cost and timely response to market changes. Appropriate business strategies and manufacturing technologies must thus be used to implement these strategic dimensions. The paradigm of Reconfigurable Manufacturing System (RMS) has been introduced to respond to this new market oriented manufacturing environment. The design of RMS allows ease of reconfiguration as it has a modular structure in terms of software and hardware. This allows ease of reconfiguration as a strategy to adapt to changing market demands. Modularity will allow the ability to integrate/remove software/hardware modules without affecting the rest of the system. RMS can therefore be quickly reconfigured according to the production requirements of new models, it can be quickly adjusted to exact capacity requirements as the market grows and products change, and it is able to integrate new technology. In this research project, real-time, open controller is designed and developed for Reconfigurable Manufacturing Tools (RMTs). RMTs are the basic building blocks for RMS. Real time and openness of the controllers for RMT would allow firstly, for the modular design of RMTs (so that RMTs can be adapted easily for changing product demands) and secondly, prompt control of RMT for diagnosability.

This project established that the new modeling mechanism--mass estimation--has a strong theoretical underpinning for prediction and data modeling. It showed that mass-based approaches have time and space complexities more favorable than existing approaches in a number of data mining tasks e.g., anomaly detection, clustering and information retrieval, and developed (i) a new density estimator based on mass, and (ii) a new generative classifier based on mass. The results have been published in top conferences and accepted for top journals.

On-line social networking has benefits but also risks to be considered when using such sites as Facebook. Internet privacy threats are a challenge that is impossible to completely mitigate on every social network, but there are steps one can take to significantly reduce the risks., Published in Army Communicator, p41-47, Spring 2011.

Seven months ago, none of us here at the School of Information Technology, Fort Gordon, Ga., knew how to develop applications for smartphones. Now we have more than 100,000 downloads of our apps in the public app stores. Such is the fast pace of technology., Published in ARMY Magazine, p23-24, Feb 2011. The original document contains color images.

Geospatial Decision Support Products (GDSPs), computer-based tools that help interpret geospatial data, are ubiquitous within current military forces. Although GDSPs are used widely, and are generally perceived as useful, there is too little empirical evidence to assess the value of GDSPs to the warfighter. Achieving a better understanding of the potential value of GDSPs while they are still in development would help to prioritize scarce intellectual, physical, and monetary resources, and would result in having the most valuable GDSPs available to the warfighter. General experimental methods are well understood and have been used recently to evaluate GDSPs during development (e.g., Powell et al., 2009). However, it can take considerable time and effort to recruit the type and number of participants required to conduct experiments with adequate statistical power. This paper describes how formative evaluations emulating experiments in terms of hypotheses, scenarios, procedures, and measures, but using only a small number of subject matter experts (SMEs), can be used effectively to inform development decisions about the potential value of GDSPs. A case study in which a formative evaluation was used to inform development decisions for the Battlespace Terrain Reasoning and Awareness-Battle Command (BTRA-BC) Tactical Spatial Objects (TSOs) is presented to illustrate the method., Presented at the International Command and Control Research and Technology Symposium (ICCRTS '10) (15th), held in Santa Monica, CA, on 22-24 Jun 2010. Prepared in cooperation with Viecore, FSD, Eatontown, NJ. The original document contains color images.

This project developed an adaptive anomaly detection system based on Isolation Forest, applicable to data stream which demands single-scan online algorithms with poly-logarithmic time and space complexities. The proposed system based on Half-Space Tree, an extension of Isolation Forest, is not only capable of detecting anomalies when the underlying concept changes gradually over time, but also capable of detecting abrupt changes in the underlying concepts. Half-Space Trees is significantly better than three existing state-of-the-art distance-based and density-based methods, in terms of detection accuracy, time complexity and memory requirement.

Results from several basic research tasks are reported. In order to respond to operations other than conventional major theater war, organizational designs that enable adaptation to changing situations are needed. In order to enable such adaptation, the supporting physical system of systems must also be adaptable and agile. Measures for assessing the adaptivity and agility of systems of systems that support the command and control functions were developed and a methodology based on executable models of architectures was designed and applied to a naval example. A second thread re-examined organization design algorithms to enhance them so that they can address cultural differences in coalition operations. A third thread used Timed Influence nets to develop and evaluate Courses of Action that would result in achieving desired effects. Model Driven Experimentation was used to develop and assess measures for evaluating the extent of to which effects based planning resulted in achieving the desired effects.

Achieving interoperability in sharing and exchanging data between health information systems requires the support of standard medical terminology. To integrate standardised terminology into information systems, there is a need to map legacy interface terminology to a reference terminology. In this study, we mapped ICPC-2 PLUS, the interface terminology developed in Australia and classified to the International Classification of Primary Care Version 2, to the SNOMED CT terminology. We have developed a series of automated mapping algorithms to assist humans to perform the mapping. The Unified Medical Language System (UMLS) metathesaurus mapping, which utilises the links between ICPC-2 PLUS and SNOMED CT terms in the UMLS library mapped 46.5% of ICPC-2 PLUS terms to SNOMED CT. Lexical mapping explored the lexical similarities between terms in these two terminologies, and mapped 60.3% of ICPC-2 PLUS terms overall. Post-coordination of remaining unmapped terms was performed, allowing one ICPC-2 PLUS term to be mapped into composition with two SNOMED CT terms, which gives an increase of about 20% in mapped terms. Overall we have mapped 80.58% of ICPC-2 PLUS terms. A manual review of the mapping shows that about 90% of string-based mappings are accurate. Unmapped terms and mismatched terms are due to the differences in the structures between these two terminologies. Also, terms contained in ICPC-2 PLUS but not in SNOMED CT caused a large proportion of failures in the mappings.

In this report, we describe a severe Mobile Ad Hoc Network (MANET) routing attack called a wormhole attack and review state-of-the-art ways to thwart wormhole attacks. In a wormhole attack, intruders tunnel the data from one end of the network to the other, leading distant network nodes to believe they are neighbours and making them communicate through the wormhole link. Unlike many other attacks on ad hoc routing, a wormhole attack can not be prevented with cryptographic solutions because intruders neither generate new, nor modify existing, packets, but rather forward existing ones., The original document contains color images.

In the modern business world, collaborative data mining becomes especially important because of the mutual benefit it brings to the collaborators. During the collaboration, each party of the collaboration needs to share its data with other parties. If the parties don't care about their data privacy, the collaboration can be easily achieved. However, if the parties don't want to disclose their private data to each other, can they still achieve the collaboration? To use the existing data mining algorithms, all parties need to send their data to a trusted central place to conduct the mining. However in situations with privacy concerns, parties may not trust anyone, including a third party. Generic solutions for any kind of secure collaborative computing exist in the literature. However, none of the proposed generic solutions is practical in handling large-scale data sets because of the prohibitive extra cost in protecting data privacy. Therefore, practical solutions need to be developed. This need underlies the rationale for our research.

A Monte Carlo simulation code has been developed that calculates the isotopes produced by spallation and propagates the products of radioactive decays for the purpose of studying activation backgrounds in space based gamma-ray telescopes. This code has been applied to study a Compton telescope consisting of two large arrays of position sensitive germanium detectors. Activation of the germanium by cosmic rays and their secondary particles is the major source of background. These simulations quantify the contributions and spectrum of each background component., Prepared in collaboration with the Universities Space Research Association and the Naval Research Laboratory, Washington, DC. Presented at the IEEE Nuclear Science Symposium, held in Albuquerque, NM, on 9-15 Nov 1997 and published in proceedings of the same, v1, p325-329, 1997. Conference Record N23-3.

Investigation of the AQ17-HCI method within the general purpose constructive-induction system AQ17 began with experimentation involving its supporting -induction system AQ15 and a target concept from the ROBOTS domain in the EMERALD system. With the aid of the diagrammatic visualization (OlAV) technique, AQ15 results from incremental training example sets were clearly displayed on the screen contrasting the learned concept against the target concept. AQ15 had searched successfully through the space of possible descriptions with its ultimate goal of inducing a positive output hypothesis that matched exactly the target concept. Investigation of the constructive-induction capabilities of AQ17-HCI proceeded with an analysis of output hypotheses from the attribute-removal module AQ17(HCI/ar). The analysis disclosed that AQ17 established the attributes occurring in the positive output hypothesis as relevant based on an attribute-utility measure. It discarded all other attributes as irrelevant. As a consequence, only relevant attributes appear in the negative output hypothesis which may lead to an overlap of the hypotheses: certain combinations of attribute values occurring in the positive output hypothesis may also occur in the negative output hypothesis. As part of the analysis, I applied a method described by Wnek, which I dubbed the altered positive output hypothesis (APOH) method and which uses what I call hypothesis-driven deduction (HOD). The APOH method yielded a final positive hypothesis which matched exactly the target concept without the incremental addition of training examples. My investigation led me to a comparison of the APOH method with Wnek's Vs* (version space with the STAR) method. When the methods initially encounter a coverage overlap by positive and negative rules, each method proceeds with a successful effort to output a final hypothesis fo, Sponsored in part by ONR grants N00014-88-K-0226, N00014-88-K-0397 and N00014-91-J-1351. Also sponsored in by DARPA.

This article presents the central ideas (as they are seen by the authors) which were discussed during a decisionmaking and knowledge-based sciences workshop held in Northern Virginia in 1985. The workshop explored research issues, and also addressed an agenda designed to take advantage of a multi-disciplinary approach. Research having a potentially high payoff in the areas of knowledge-based support of human information processing, judgement, and choice is described. This article also serves as the conference's final report.

This research note summarizes research and design needs in the area of computer based intelligence support, particularly the command and control area. It is argued that an integrating approach is needed. A discussion of knowledge representation emphasizes this, and an approach that accommodates both probabilistic and logical support, and which is able to cope with several types of imperfect information is described.

Many authors have used distribution functions related to the exponential staging formulation introduced by Erlang (see Brockmeyer et al., 1948), with primary modifications over the years by Jensen (1954) and Cox (1955) . Much of the current popularity of such distributions is due to the work of Neuts and colleagues on the so-called phase-type family (see, e.g., Neuts, 1975 a,b and 1981), exploiting relationships to the theory of Markov chains and putting the theory to effective computational use in a large variety of stochastic models. For purposes of clarity, we note the following relationships between these classes of cumulative distribution functions (CDFs). Essentially, all of this stems from Erlang's early idea of modeling a duration or lifetime as a sum of independent and identical exponential stages. Much later, Jensen (1954) generalized Erlang's device to allow the stages to have non-identical CDFs, and indeed recognized the natural connection between Erlang's method of stages and absorption-time distributions of finite Markov chains. (KR)