International Journal of Agent Technologies and Systems, 2016
Thisarticledescribeschangingcustomerdemandsrequirethatenterprisesmobilizetheirresources to quickl... more Thisarticledescribeschangingcustomerdemandsrequirethatenterprisesmobilizetheirresources to quickly develop a suitable product. This is achievable if competing enterprises collaborate to delivertheproduct.Eachofthembringstheirexpertiseintothecollaboration.Thiscollaboration whereeachenterprisebringsinitscorecompetencyisreferredtoasavirtualenterprise(VE).A constructionprojectisimplementedbyateamofprofessionalsandanallianceofcompaniesthatis formedbyconsultantswhoevaluatecontractorsforspecificprojecttasks.Partnerscanberepresented asmultipleagents.Priorevidenceofmulti-agentsystem(MAS)modelthatfacilitatesformationof VEsislacking.VEMASontologyhasbeendesignedandusedinagentinteractions.Themodel canbeusedinevaluationandselectionprocessofpartners.Delegationoftheprocesstothemodel, givespartnerstimetoimplementthetasks.Partnerevaluationandselectionproblemforbuilding constructionprojectsissolvableifpragmaticscientificapproachesareemployedwithappropriate mathematicalmodels.ThisarticleproposedaVEmodelforevaluatingandselectingrightpartnersfor buildingconstructionprojects.Themodelwasusedtodemonstratethechoiceofthemostpreferred partner.Researchershavenotevaluatedthismodelbutproposethatonceinplace,itcanevaluated againstmanualselectionofpotentialpartnersusingsimilarparametersbyexaminingthecloseness oftheoutput.
International journal of machine learning and applications, May 23, 2013
Many scholars are interested in improving e-learning in order to provide easy access to education... more Many scholars are interested in improving e-learning in order to provide easy access to educational materials. There is, however, the need to incorporate the ability to classify learners into these learning systems. Learner classification is used adaptively to provide relevant information for the various categories of learners. There is also a need for learning to continue, whether learners are on-or off-line. In many parts of the world, especially in the developing world, most people do not have reliable continuous internet connections. We tested an Adaptive e-Learning Model prototype that implements an adaptive presentation of course content under conditions of intermittent Internet connections. This prototype was tested in February 2011 on undergraduate students studying a database systems course. This study found out that it is possible to have models that can adapt to characteristics such as the learner's level of knowledge and that it is possible for learners to be able to study under both on-and off-line modes through adaptation.
Recently, demand for database programming specialists has greatly increased in Kenya. These profe... more Recently, demand for database programming specialists has greatly increased in Kenya. These professionals play a key role in the computing and software development industries. Although database programming skills are key fundamentals for learners in computing disciplines, skills mastery by students is still not easy. For these reasons, this study establishes an adaptive remedial learning model to assist learners in their quest of gaining skills online. The proposed solution adopts the use of fuzzy logic theory to create an appropriate learning path based on the learners’ prior concepts miscomprehensions. This technique selects a suitable remedial materials for learners after constructing a learning path based on the learners’ preference. After evaluation of the model through conducting several experiments, it is proposed that it can be used to offer a comprehensive and stable remedial learning environment for any LMS. Analysis of the model by learners confirm that it has achieved the effects of remedial and adaptive learning.
International Journal of Modeling and Optimization
The banking industry has seen a revolutionary moment in adopting mobile banking technology in the... more The banking industry has seen a revolutionary moment in adopting mobile banking technology in the last few years. The technology has contributed to many customers keeping off the banking hall for basic and routine transactions. This has largely been attributed to the convenience of mobile banking, including time management and increased privacy. However, there are still times when a bank conducts large-scale service provision for many people. For example, a payment agency contracted by the government to disburse funds to a section of the citizens may require that money be channeled only through a particular bank. If the period for such a requirement is limited, there is likely to be pressure on both the bank and the would-be customers because of long queues. For cases like that, there is a need to simulate possible scenarios and plan to avoid delays and frustrations for both parties. This work proposes a queue simulation model that can be used to forecast the number of bank staff th...
Zenodo (CERN European Organization for Nuclear Research), Jan 9, 2023
Bandits with Knapsacks (BwK) is a Multi-Armed Bandit (MAB) problem under supply/budget constraint... more Bandits with Knapsacks (BwK) is a Multi-Armed Bandit (MAB) problem under supply/budget constraints. Risk scoring is a typical limited-resource problem and as such can be modeled as a BwK problem. This paper tries to solve the triple problem in risk scoring accuracy, fairness, and auditability by proposing FuzzyBwK application in risk scoring. Theoretical assessment of FuzzyBwK is made to establish whether the regret function would perform better than StochasticBwK and AdversarialBwK functions. An empirical experiment is then set with secondary data (Australian and German credit data) and primary data (Kakamega insurance data) to determine whether the algorithm proposed would be fit for the proposed problem. The results show that the proposed algorithm has optimal regret function and from the empirical test, the algorithm satisfies the test of accuracy, fairness and auditability.
International Journal of Computer Trends and Technology
Small-scale farmers suffer unfairness during credit risk scoring. This arises from the fact that ... more Small-scale farmers suffer unfairness during credit risk scoring. This arises from the fact that scoring done using computer machine-learning algorithms has an inherent bias, otherwise called algorithm bias. The data that the small-scale farmers present is another source of bias. This paper explores these data types to bring out the specific challenges with the data and how the same can be remedied. The research findings show that of the possible 23 data types lenders ask from farmers, 14 are regarded as important. Out of these 14, 7 are commonly unavailable while the remaining 7 are not, introducing missing data records. The findings also show that other than the personal/behavioral data that the loan-seeker provides, where the lender asks for historical or environmental data, there is room for the loan-seeker to provide misleading information. This paper proposes 14 data types that can improve the quality of credit risk scoring. The study further proposes using the Internet of things and blockchain to source the environmental and historical data to improve the availability of the missing and outlier challenge in data.
2021 International Conference on Electrical, Computer and Energy Technologies (ICECET), 2021
Research on Influence Maximization has gained a lot of attention in the recent past. Part of the ... more Research on Influence Maximization has gained a lot of attention in the recent past. Part of the reason for this is that influence maximization has applications in commercially attractive areas such as word of mouth marketing. A majority of works in influence maximization have relied on information diffusion models that largely ignore the structural properties of the social network. The problem with this is that important attributes of user relationships which are necessary for approxi-mating influence maximization are ignored. Parameters such as Homophily and Topological Overlap are crucial determinants of the level of influence that a user enjoys in the network. This work approximates global influence power of a user by first, considering user interactions, homophily and topological overlap as determinants of node to node relationship strength on a dynamic social graph. Then secondly, computes a global score of influence for each user. We then apply a novel algorithm that approximates influence spread for each influential user. The seed set is built by identifying the most influential users at specific time instances as the social graph evolves.
Journal of Encapsulation and Adsorption Sciences, 2020
Many science-based institutions in most developing countries use heavy metal containing salts in ... more Many science-based institutions in most developing countries use heavy metal containing salts in practical teaching sessions. The commonly used chemicals are the salts of lead (II) and copper (II) and the wastes generated end up into the environment when untreated. Thus, a study was done to remove lead (II) and copper (II) ions from mono synthetic aqueous solution using bio-char from Ficus natalensis fruits (FNF). This was done at varied pH, contact time, temperature, bio-char dosage level, salinity and metal ion concentration using the batch approach. The residual metal concentrations were determined using the atomic absorption spectrophotometer. The optimum pH for the adsorption of copper (II) and lead (II) ions was found to be 4.0 and 5.0 respectively. The maximum percentage adsorption of copper (II) and lead (II) by the FNF bio-char was established at 60 minutes contact time, 47.5˚C and 0.4 g adsorbent dose. Increase in the metal ion concentration and the presence of interfering ions in the aqueous solution lead to decrease in the percentage adsorption. The highest adsorption capacity was found to be 161.29 mg/g and 1250 mg/g for copper (II) and lead (II) ions respectively. The thermodynamic parameters indicated the feasibility of the adsorption of copper (II) and lead (II) on the bio-char of FNF. Thus, bio-char from FNF may be used as an adsorbent in waste management where copper (II) and lead (II) ions are present at a concentration range of between 5 and 100 mg/l.
All service providers strive to attain superior service quality (SQ), since the demand for provis... more All service providers strive to attain superior service quality (SQ), since the demand for provision of high quality customer focused services is constantly rising. In the healthcare sector, it becomes of utmost importance to understand consumer expectations and their needs to effectively remain competitive in the market. Increase in the per capita income of clients has made them increase their aspiration level to demand enhanced quality services. Currently, hospitals in Nairobi City County (NCC) in Kenya, do not offer services to patients based on their aspirations and in the process, they lose clients. There is need for hospital administrators to determine how patients perceive healthcare quality service, and employ techniques that match their expectations. This is probable by using patients to evaluate their services based on identified criteria. Human evaluations are normally imprecise thus uncertainty aspects should be integrated in the evaluation process. Techniques employed s...
Group decision making takes place in almost all domains. In building construction domain, a team ... more Group decision making takes place in almost all domains. In building construction domain, a team of contractors with disparate specializations collaborate. Little research has been done to propose group decision making technique for this domain. As such, specific teams’ competitiveness enhancements are minimal as it takes more time for individual evaluators to choose the right partners. Qualitative and quantitative methods were used. Themes and categorizations were based on deductive approach. Subsequently, Group Fuzzy Analytical Hierarchy Process (GFAHP), Multi-Criteria Decision Making (MCDM) algorithm, was designed and applied. It uses all evaluation criteria unlike Fuzzy AHP (FAHP) which excludes some criteria that are assigned zero weights. GFAHP reduces the number of pairwise comparisons required when a large number of attributes are to be compared. Validation of the technique carried out by five case studies, show that GFAHP is approximately 98.7% accurate in the selection of ...
Consumer lending is an exercise where risk scoring takes the form of a typical decision making pr... more Consumer lending is an exercise where risk scoring takes the form of a typical decision making problem. For smallholder farmers, the credit scoring becomes specifically more challenging with the data gaps and outliers in data. Added to that, the process must be as cost-effective as possible while providing as accurate results as possible. This paper uses data obtained from smallholder farmers in Kakamega County in Kenya to set up an experiment of credit scoring as a Bandit with Knapsack problem with Fuzzy Unordered Rule Induction Algorithm (FURIA) being used as the exploit-explore algorithm and Fuzzy Analytical Hierarchical Process (FAHP) used to determine the ranking and consistency of the FURIA rules. The experiment returns a consistency ratio of 0.000529 which is significantly less than the 0.10 threshold. In this regard, the paper proposes the use of FURIA to reduce the regret in Bandit with Knapsack (BwK) as a technique for smallholder credit scoring.
In consumer loans, where lenders deal with masses, use of algorithms to classify borrowers is fas... more In consumer loans, where lenders deal with masses, use of algorithms to classify borrowers is fast catching up. Classification based on predictive models tend, to adversely affect borrowers. In this paper, we study the extent to which various algorithms disenfranchise borrowers lying on the boundaries of decision making. In the study, the data used for loan appraisal, and decisions made by the lenders are subjected to a set of select algorithms. The bias suffered by borrowers in each case is determined using mean absolute error (MAE) and relative absolute error (RAE). The results show that FURIA has the least bias with the MAE of 0.2662 and 0.1501 and RAE of 64.19% and 30.31% for the German and Australian data sets respectively. Consequently, FURIA is modified to remove the hard boundaries which results in even lower MAE of 0.2535 and 0.1264 and RAE of 64.14% and 27.73% for the German and Australian data sets respectively.
International Journal of Agent Technologies and Systems, 2016
Thisarticledescribeschangingcustomerdemandsrequirethatenterprisesmobilizetheirresources to quickl... more Thisarticledescribeschangingcustomerdemandsrequirethatenterprisesmobilizetheirresources to quickly develop a suitable product. This is achievable if competing enterprises collaborate to delivertheproduct.Eachofthembringstheirexpertiseintothecollaboration.Thiscollaboration whereeachenterprisebringsinitscorecompetencyisreferredtoasavirtualenterprise(VE).A constructionprojectisimplementedbyateamofprofessionalsandanallianceofcompaniesthatis formedbyconsultantswhoevaluatecontractorsforspecificprojecttasks.Partnerscanberepresented asmultipleagents.Priorevidenceofmulti-agentsystem(MAS)modelthatfacilitatesformationof VEsislacking.VEMASontologyhasbeendesignedandusedinagentinteractions.Themodel canbeusedinevaluationandselectionprocessofpartners.Delegationoftheprocesstothemodel, givespartnerstimetoimplementthetasks.Partnerevaluationandselectionproblemforbuilding constructionprojectsissolvableifpragmaticscientificapproachesareemployedwithappropriate mathematicalmodels.ThisarticleproposedaVEmodelforevaluatingandselectingrightpartnersfor buildingconstructionprojects.Themodelwasusedtodemonstratethechoiceofthemostpreferred partner.Researchershavenotevaluatedthismodelbutproposethatonceinplace,itcanevaluated againstmanualselectionofpotentialpartnersusingsimilarparametersbyexaminingthecloseness oftheoutput.
International journal of machine learning and applications, May 23, 2013
Many scholars are interested in improving e-learning in order to provide easy access to education... more Many scholars are interested in improving e-learning in order to provide easy access to educational materials. There is, however, the need to incorporate the ability to classify learners into these learning systems. Learner classification is used adaptively to provide relevant information for the various categories of learners. There is also a need for learning to continue, whether learners are on-or off-line. In many parts of the world, especially in the developing world, most people do not have reliable continuous internet connections. We tested an Adaptive e-Learning Model prototype that implements an adaptive presentation of course content under conditions of intermittent Internet connections. This prototype was tested in February 2011 on undergraduate students studying a database systems course. This study found out that it is possible to have models that can adapt to characteristics such as the learner's level of knowledge and that it is possible for learners to be able to study under both on-and off-line modes through adaptation.
Recently, demand for database programming specialists has greatly increased in Kenya. These profe... more Recently, demand for database programming specialists has greatly increased in Kenya. These professionals play a key role in the computing and software development industries. Although database programming skills are key fundamentals for learners in computing disciplines, skills mastery by students is still not easy. For these reasons, this study establishes an adaptive remedial learning model to assist learners in their quest of gaining skills online. The proposed solution adopts the use of fuzzy logic theory to create an appropriate learning path based on the learners’ prior concepts miscomprehensions. This technique selects a suitable remedial materials for learners after constructing a learning path based on the learners’ preference. After evaluation of the model through conducting several experiments, it is proposed that it can be used to offer a comprehensive and stable remedial learning environment for any LMS. Analysis of the model by learners confirm that it has achieved the effects of remedial and adaptive learning.
International Journal of Modeling and Optimization
The banking industry has seen a revolutionary moment in adopting mobile banking technology in the... more The banking industry has seen a revolutionary moment in adopting mobile banking technology in the last few years. The technology has contributed to many customers keeping off the banking hall for basic and routine transactions. This has largely been attributed to the convenience of mobile banking, including time management and increased privacy. However, there are still times when a bank conducts large-scale service provision for many people. For example, a payment agency contracted by the government to disburse funds to a section of the citizens may require that money be channeled only through a particular bank. If the period for such a requirement is limited, there is likely to be pressure on both the bank and the would-be customers because of long queues. For cases like that, there is a need to simulate possible scenarios and plan to avoid delays and frustrations for both parties. This work proposes a queue simulation model that can be used to forecast the number of bank staff th...
Zenodo (CERN European Organization for Nuclear Research), Jan 9, 2023
Bandits with Knapsacks (BwK) is a Multi-Armed Bandit (MAB) problem under supply/budget constraint... more Bandits with Knapsacks (BwK) is a Multi-Armed Bandit (MAB) problem under supply/budget constraints. Risk scoring is a typical limited-resource problem and as such can be modeled as a BwK problem. This paper tries to solve the triple problem in risk scoring accuracy, fairness, and auditability by proposing FuzzyBwK application in risk scoring. Theoretical assessment of FuzzyBwK is made to establish whether the regret function would perform better than StochasticBwK and AdversarialBwK functions. An empirical experiment is then set with secondary data (Australian and German credit data) and primary data (Kakamega insurance data) to determine whether the algorithm proposed would be fit for the proposed problem. The results show that the proposed algorithm has optimal regret function and from the empirical test, the algorithm satisfies the test of accuracy, fairness and auditability.
International Journal of Computer Trends and Technology
Small-scale farmers suffer unfairness during credit risk scoring. This arises from the fact that ... more Small-scale farmers suffer unfairness during credit risk scoring. This arises from the fact that scoring done using computer machine-learning algorithms has an inherent bias, otherwise called algorithm bias. The data that the small-scale farmers present is another source of bias. This paper explores these data types to bring out the specific challenges with the data and how the same can be remedied. The research findings show that of the possible 23 data types lenders ask from farmers, 14 are regarded as important. Out of these 14, 7 are commonly unavailable while the remaining 7 are not, introducing missing data records. The findings also show that other than the personal/behavioral data that the loan-seeker provides, where the lender asks for historical or environmental data, there is room for the loan-seeker to provide misleading information. This paper proposes 14 data types that can improve the quality of credit risk scoring. The study further proposes using the Internet of things and blockchain to source the environmental and historical data to improve the availability of the missing and outlier challenge in data.
2021 International Conference on Electrical, Computer and Energy Technologies (ICECET), 2021
Research on Influence Maximization has gained a lot of attention in the recent past. Part of the ... more Research on Influence Maximization has gained a lot of attention in the recent past. Part of the reason for this is that influence maximization has applications in commercially attractive areas such as word of mouth marketing. A majority of works in influence maximization have relied on information diffusion models that largely ignore the structural properties of the social network. The problem with this is that important attributes of user relationships which are necessary for approxi-mating influence maximization are ignored. Parameters such as Homophily and Topological Overlap are crucial determinants of the level of influence that a user enjoys in the network. This work approximates global influence power of a user by first, considering user interactions, homophily and topological overlap as determinants of node to node relationship strength on a dynamic social graph. Then secondly, computes a global score of influence for each user. We then apply a novel algorithm that approximates influence spread for each influential user. The seed set is built by identifying the most influential users at specific time instances as the social graph evolves.
Journal of Encapsulation and Adsorption Sciences, 2020
Many science-based institutions in most developing countries use heavy metal containing salts in ... more Many science-based institutions in most developing countries use heavy metal containing salts in practical teaching sessions. The commonly used chemicals are the salts of lead (II) and copper (II) and the wastes generated end up into the environment when untreated. Thus, a study was done to remove lead (II) and copper (II) ions from mono synthetic aqueous solution using bio-char from Ficus natalensis fruits (FNF). This was done at varied pH, contact time, temperature, bio-char dosage level, salinity and metal ion concentration using the batch approach. The residual metal concentrations were determined using the atomic absorption spectrophotometer. The optimum pH for the adsorption of copper (II) and lead (II) ions was found to be 4.0 and 5.0 respectively. The maximum percentage adsorption of copper (II) and lead (II) by the FNF bio-char was established at 60 minutes contact time, 47.5˚C and 0.4 g adsorbent dose. Increase in the metal ion concentration and the presence of interfering ions in the aqueous solution lead to decrease in the percentage adsorption. The highest adsorption capacity was found to be 161.29 mg/g and 1250 mg/g for copper (II) and lead (II) ions respectively. The thermodynamic parameters indicated the feasibility of the adsorption of copper (II) and lead (II) on the bio-char of FNF. Thus, bio-char from FNF may be used as an adsorbent in waste management where copper (II) and lead (II) ions are present at a concentration range of between 5 and 100 mg/l.
All service providers strive to attain superior service quality (SQ), since the demand for provis... more All service providers strive to attain superior service quality (SQ), since the demand for provision of high quality customer focused services is constantly rising. In the healthcare sector, it becomes of utmost importance to understand consumer expectations and their needs to effectively remain competitive in the market. Increase in the per capita income of clients has made them increase their aspiration level to demand enhanced quality services. Currently, hospitals in Nairobi City County (NCC) in Kenya, do not offer services to patients based on their aspirations and in the process, they lose clients. There is need for hospital administrators to determine how patients perceive healthcare quality service, and employ techniques that match their expectations. This is probable by using patients to evaluate their services based on identified criteria. Human evaluations are normally imprecise thus uncertainty aspects should be integrated in the evaluation process. Techniques employed s...
Group decision making takes place in almost all domains. In building construction domain, a team ... more Group decision making takes place in almost all domains. In building construction domain, a team of contractors with disparate specializations collaborate. Little research has been done to propose group decision making technique for this domain. As such, specific teams’ competitiveness enhancements are minimal as it takes more time for individual evaluators to choose the right partners. Qualitative and quantitative methods were used. Themes and categorizations were based on deductive approach. Subsequently, Group Fuzzy Analytical Hierarchy Process (GFAHP), Multi-Criteria Decision Making (MCDM) algorithm, was designed and applied. It uses all evaluation criteria unlike Fuzzy AHP (FAHP) which excludes some criteria that are assigned zero weights. GFAHP reduces the number of pairwise comparisons required when a large number of attributes are to be compared. Validation of the technique carried out by five case studies, show that GFAHP is approximately 98.7% accurate in the selection of ...
Consumer lending is an exercise where risk scoring takes the form of a typical decision making pr... more Consumer lending is an exercise where risk scoring takes the form of a typical decision making problem. For smallholder farmers, the credit scoring becomes specifically more challenging with the data gaps and outliers in data. Added to that, the process must be as cost-effective as possible while providing as accurate results as possible. This paper uses data obtained from smallholder farmers in Kakamega County in Kenya to set up an experiment of credit scoring as a Bandit with Knapsack problem with Fuzzy Unordered Rule Induction Algorithm (FURIA) being used as the exploit-explore algorithm and Fuzzy Analytical Hierarchical Process (FAHP) used to determine the ranking and consistency of the FURIA rules. The experiment returns a consistency ratio of 0.000529 which is significantly less than the 0.10 threshold. In this regard, the paper proposes the use of FURIA to reduce the regret in Bandit with Knapsack (BwK) as a technique for smallholder credit scoring.
In consumer loans, where lenders deal with masses, use of algorithms to classify borrowers is fas... more In consumer loans, where lenders deal with masses, use of algorithms to classify borrowers is fast catching up. Classification based on predictive models tend, to adversely affect borrowers. In this paper, we study the extent to which various algorithms disenfranchise borrowers lying on the boundaries of decision making. In the study, the data used for loan appraisal, and decisions made by the lenders are subjected to a set of select algorithms. The bias suffered by borrowers in each case is determined using mean absolute error (MAE) and relative absolute error (RAE). The results show that FURIA has the least bias with the MAE of 0.2662 and 0.1501 and RAE of 64.19% and 30.31% for the German and Australian data sets respectively. Consequently, FURIA is modified to remove the hard boundaries which results in even lower MAE of 0.2535 and 0.1264 and RAE of 64.14% and 27.73% for the German and Australian data sets respectively.
Uploads
Papers by George Musumba