Malaria is a mosquito-borne infectious disease caused by protists (a form of microorganism) of the Plasmodium genus in humans and other animals. Malaria is a leading worldwide cause of morbidity and mortality. According to WHO the estimated value of malaria cases in 2019 was 229 million worldwide, with children under the age of 5 years having 67% (274,000) and being the most vulnerable group affected by malaria. Nigeria Demographics and health survey (NDHS) has a repository for data that can be used to predict malaria disease outbreak using machine learning techniques, from the literature reviewed no research has been carried out using machine learning technique to model the prediction of malaria outbreak using malaria incidence data from southwest Nigeria. This Research work used 5 supervised machine learning techniques to model the outbreak of malaria using meteorological and malaria incidence data of collected from 2010 - 2020, the machine learning techniques that was used are, Naive Bayes, Support Vector, Linear Regression, Logistic Regression, and K-Nearest Neighbor. The research was carried out using Scikit-learn Library that was imported into Anaconda IDE, the programming language used was Python programming language, The result of the research shows that Naive Bayes has the best accuracy for both testing and training with average accuracy of 79.1% and therefore is the best prediction model that can be used for predicting malaria incidence outbreak using the data set used in this research, Support Vector machine (SVM) is the second best prediction model that can be used for predicting malaria incidence outbreak for both testing and training data with average accuracy of 75.45%, followed by K-Nearest Neighbor with average accuracy of 70.8%, followed by Logistic Regression prediction model which has an average accuracy of 68%, based on this research work it is not advisable to use Linear Regression prediction model for predicting malaria incidence outbreak because it has an average accuracy of 26.05%.

Artificial Intelligence, Machine Learning, pre-processing, Prediction, Malaria, Support Vector machine, Naive Bayes, Logistic Regression

The role of credit risk’s models for financial institutions and the economy at large in making lending decisions cannot be overemphasised. However, there is many doubts if these scientific models used to predict credit risks are applied by financial institutions in developing countries going by huge nonperforming loans recorded banks in these nations annually. This research examined the awareness and application of machine learning models for credit risk predictions among financial institutions in developing countries with focus on Nigeria. Structured questionnaires were developed for data collection. The statistical package for Social Sciences version 17 (SPSS 17) was used to calculate the mean responses. Mean (X ̅) was used to answer the research questions. The result shows a very low awareness on machine learning techniques while its application for credit risk prediction is completely not in place among the financial institutions in developing countries. The study therefore recommends the deployment and application of machine learning techniques for credit risk prediction in developing countries which will serve as an objective basis for assessing loan applicants to replace the judgmental system which is mainly a subjective decision that is bond to errors and human interference.

Machine Learning, Credit Risk, Artificial Intelligence, models and technique

Attendance taking is one of the important activities during lectures in reputable academic institutions as it can be used for important decision making. There are number of methods for taking attendance. The method of taking attendance can make the exercise to be boring, waste of time and even dangerous. The use of manual and biometrics attendance system has negative impacts. In recent time, with the use of internet and mobile devices, attendance system can be carried out easily without delay or danger to human. In this study, the use of Quick Response code for attendance taking has been developed for Ajayi Crowther University. Various information about attendance system can be obtained easily with the use of the proposed system.

Attendance, lectures, biometrics, manual, Quick Response code

E-government is being adopted by many governments in recent times as a means for easier and faster way for the provision of government services to the citizenry. Despite the known advantages of e-government, literature show that some e-government implementations fail to meet the maximum quality standards. Poor quality e-government services fail to provide the needed quality of service to the citizenry. Like any other web platform, it is imperative to evaluate e-government platforms to ensure maximised quality of service for the benefit of the citizenry. This study evaluates five (5) e-government applications in Ghana based on the ISO 9126 quality model. Results from the study shows that all the five (5) evaluated applications are highly functional and portable. However, there were issues with the reliability test since some of the applications performed poorly when they were subjected to stress. It is recommended that the reliability of those applications be thoroughly reviewed.

E-government, software quality, web applications, mathematical model, ISO quality model

This paper proposes an Embedded Arduino Automatic Irrigation-Control System (EAAICS), which monitors and maintains the desired soil moisture content via automatic watering using embedded technology. The methodology employed in this paper is solely based on the integration of both hardware and software components to design and implement an App-controlled irrigation system, which solely depends on the soil moisture content. Input to the system was derived from soil samples with the aid of a soil moisture sensor and a microcontroller controlled the rate of flow of water from an electric water pump. A prototype was tested using a flower pot containing an ornamental plant, which was watered adequately through the mobile application and it grew to become a beautiful and blossom plant after forty days. The prototype showed that it would really help bring more yield from crop production if introduced into many farms regardless of the size. To perceive the ease of system’s usefulness, it was later introduced to small and large-scale farmers to aid their farming activities and evaluated with the Technology Accepted Model (TAM). The results yielded a high acceptance rate by the respondents, which in turn signifies the importance of introducing embedded technology in automating irrigation system for better productivity.

Irrigation, Soil moisture, Arduino Technology, Embedded system Application, Controller

In this paper, an improved exponential type estimator for estimating the population mean is proposed under simple random sampling scheme. The proposed estimator was obtained by combination of conventional product and exponential-type ratio estimators with aim of obtaining estimator with higher efficiency. The bias and mean squared error (MSE) of the proposed estimator were obtained up to the first order of approximation using binomial and exponential expansion techniques and the optimum value of the unknown constant of the estimator was derived by means of partially differentiating the mean squared error and equating to zero. Also, the conditions under which the proposed estimator is more efficient than the conventional estimators in the literature are established. An empirical study was carried out to support the fact that the proposed estimator is better than the existing ones, as the proposed estimator has a minimum mean squared error at the optimum value of the unknown constant and has higher percentage relative efficiency (PRE). This implies that the proposed estimator is more efficient than the conventional product and exponential-type ratio estimators considered in the study.

Simple random sampling, Auxiliary information, Mean squared error, Efficiency

The exponential rate of increase in Nigeria population without complementary increase in science and social economic indicators not only lead to suffering from natural environmental curse, but more importantly the loss of renewable natural resources beyond sustainable limits and general environmental degradation. Soil degradation, rapid deforestation, desertification, and others are some of the current environmental problems facing Nigeria. Many efforts have been made on environmental degradation by both the government and researchers, but most of which prove abortive in finding everlasting solution to this problem. Therefore, the focus of this work is to adopt the Vector Autoregressive Model to different variables that affect environmental quality. The relationships among forest area, population, family planning, agricultural area and unemployment rate of Nigeria were examined between 1990 and 2018. The data for the study were obtained from the National Bureau of Statistics (NBS) and World Development Indicator (WDI), World Bank. The unit root test was conducted to examine the stationary level of the selected variables. Results from augmented Dickey Fuller test showed that family planning, forest area, unemployment, and agricultural area were stationary at difference of two, while population was stationary at difference of three. Information criteria were carried out to determine the lag length of the Vector Autoregressive Model and the maximum lag length for the real-life data was set to three. Correlation analysis was examined to show the level of relationship between the variables considered. The Pearson statistic showed a negative bidirectional relationship between forest area and population rate, agricultural rate and unemployment rate while a positive directional trend was experienced between forest area and family planning. Granger causality test showed that each of the variables has impact on one another. Results showed that population, family planning, agricultural area and unemployment rate significantly influence the rate of forest area of Nigeria. This study recommends that family planning should be embraced by the masses to reduce population growth in order to curb environmental degradation. Government should create job to prevent people from finding life sustenance in forest and more importantly land use act should be strictly implemented on forest area and agricultural area as there must be specification on cultivated land from forest area.

Vector autoregressive model, Population, Forest area, Unemployment, Family planning

Computer Based Test (CBT) system has become a widely used tool for assessing examinee capability in examinations which is not limited to students but also to job seekers because of its instant delivery of results. However, most of the existing work studied a number of events that contributed to the adoption of CBT but little work had been done on the study of correlation between events which made it impossible to know the effect of certain event over the others. Hence, this studied the inter-relationship between events and how the relationship impacts positive changes on the CBT system in tertiary institutions using Cross-impact method. Questionnaires were administered on experts in three selected tertiary institutions in Oyo state. The questionnaire was structured to collate the opinions of experts on the probabilities of single occurrence and conditional occurrence of Examination Policy (EP), Availability of Software and Hardware (SH), Lecturers Acceptance (LA), School Management Commitments (MC) and Students Performance (SP) which were the five major relevant events considered for the adoption of Computer based testing system in tertiary institutions. The data obtained through the questionnaires were analyzed to derive the Initial Probability and Conditional Probability which constitute the Cross-Impact Probability matrix for the occurrence of considered events. Sensitivity testing was performed on each event to determine the individual impact on others.The results of sensitivity testing of the effect of EP on others showed the significant changes of 8, 11, 4 and 10% for SH, LA, MC and SP, respectively. Also, the result of sensitivity testing of the effect of SH on others experienced the significant changes of 9, 14, 6 and 13% for EP, LA, MC and SP, respectively. The result of sensitivity testing of the effect of LA on others experienced the significant changes of 14, 16, 12 and 19% for EP, SH, MC and SP, respectively. The study showed that all five events were important and had impact on one another but lecturers’ acceptance has the highest impact while school management commitment has the least impact for the effective adoption of CBT in tertiary institutions. Thus, this will serve as baseline information for intending institutions in adopting CBT system in assessing their students’ capabilities.

Computer Based Test, Cross-Impact Method, Initial Probability, Conditional Probability, Sensitivity Testing

Human-computer interaction artificial intelligent system has become more necessary in the African region especially in Nigeria. Local innovative capabilities experience challenges with awareness, non-functionality of the system, and local professional teams carrying out local innovations. However, current human-computer interaction (HCI) guidelines are limited in their applicability to local needs. The paper aims to improve the interactions between users and designed technologies through the user interface. The objectives are to promote usability and maintain the HCIAI system (robot) efficiently; to create awareness of the functionality of HCI in our daily use; and to increase local innovative capabilities and develop the fundamental concepts within our environment. This paper describes an intuitive interaction process with specific attributes allowing interactions between humans and the intelligent system as well as eliciting target experience and making it more appealing to user interface designers as a tool.

Innovative, Human-computer interaction, Artificial intelligence, environment, user interface designer, local innovative

: The human race has been at the edge of the COVID-19 pandemic since the start of 2020. While the disease is easily transmissible, a large proportion of the people affected are recovering. Most recovered patients do not suffer COVID-19 death, even though they have been observing for a long time. In the sense of survival analysis, they can be viewed as long term survivors (cured population). In this study, we present some statistical methods for estimating the cure fraction in Kosovo of COVID-19 patients. Proportional hazards Mixture cure model is used to estimate the fraction of cure and the effect of gender and age covariations on lifetime. For this analysis the data available on the https://raw.githubusercontent.com/owid/covid-19-data/master/public/data/owid-covid-data.xlsx' website is used. The result revealed that the covariates, diabetes prevalence and hospital beds per thousand have highly statistically significant coefficients, while others, that is stringent index, total cases, gdp per capita (economic variable), respondent’s age, handwashing facilities are not statistically significant, implying that these variables are not really contributing to the hazard ratio of covid-19 incidence.

GDP per capita, Covid 19, respondent’s age, handwashing facilities

Geo-fencing is a virtual perimeter in a geographical area that uses location-based services as a boundary for an area. To provide better accuracy and improving tracking of travellers using an optimized virtual fencing (geo-fence) to detect unusual user locations. This work presents a proposed technique for testing the possibilities of merging a geo-fencing technique and Kalman Filter to provide more accurate data to the smartphone users. The location monitoring system is an LBS (Location Based Services) system which utilizes the GPS found on a smartphone. This location is subjected to error variance using Kalman filter and then sent to the server. The server will display the location of the traveller’s whereabouts on a map that can be accessed via the website or next of kins smartphone. The use of geo-fencing would limit traveller supervision areas and notification will appear to user in the form of a message or alarm via their smartphone device should in case the traveller leaves the Geo-fence area. The proposed scheme presents a model that allows family members to directly monitor the whereabouts of loved ones using the incorporation of Geofencing technology, Kalman filter, GPS, and SMS (Short Messages Services).

Geo-fence, tracking, Location Based Services, Smartphone, coordinates, Kalman Filter

Mathematical model for the control of Ebola Virus by Quarantine Technique method have been developed. The model is first order non-linear differential equation, in which the model has been divided into six compartments; exposed individual (SH). Exposed Vector (SV), individual with infection (IH), Quarantine individual (QH), Recovered individual (RH) and vector with the disease (IV). The Equilibrium State were Obtained and their Stabilities were analyzed by using A domain Decomposition Method (ADM). The result shows that when treatment rate is high, population of the infected human and quarantine human will be reduced. However, early detection of infected individual as well as treatment in time led to the reduction of Ebola Virus transmission in the Population respectively.

Ebola-Virus, Mathematical Model, Quarantine-Technique and ADM

The process of machine learning has been useful in finding solution to issues of getting important, accurate and meaningful information. This paper provides machines with the abilities of collecting data using systems like humans and processing the data using machine learning techniques for predictions and arriving at accurate decisions the same level as humans. This paper uses two types of machine learning techniques. They techniques considered in this research are Navies Bayes and K-means clustering techniques. The confusion matrix was introduced to test the performance of the machine learning algorithms. At the end of this research, a more accurate and efficient technique would be obtained for providing insights, making reliable predictions, and for accurate decision making process.

Human senses, computational methods, predictions, classification methods, confusion matrix

There is currently no study that focuses solely on the causal relationship between Crop Production Index (CPI) and Permanent Cropland (PCL) in developed and developing countries, especially in Nigeria. However, understanding the causal relationship between CPI and PCL is crucial to both food security and economic growth of any nation. In this paper, we investigated the causal relationship between CPI and PCL time series variables using unrestricted Vector Autoregressive (VAR) modeling techniques. Pre-examination of the CPI and PCL time series data extracted from the repository of World Bank showed that these two series were not only difference stationary series of order one {I(1)} but are also not cointegrated. The results of optimal lag length confirmed that VAR (3) model best fitted the data. The findings showed that Nigeria’s crop production index is predictable by Nigeria’s permanent cropland and vice versa.

Crop Production Index, Permanent Cropland, Unrestrited VAR, Optimal Lag Length, Cointegration, Nigeria

Cervical cancer related mortality in developing countries like Nigeria is alarming and coupled with the incessant industrial dispute in the medical industry, early detection and treatment of the deadly disease is arduous hence the mortality rate. In this paper, we develop a framework of a knowledge-based diagnostic support for cervical cancer using the instrumentality of predictive analytics by deploying Iris dataset for the training of four base learners. This is aimed at presenting a proactive measure towards early detection and diagnosis of the menace via a web-based use case. Experimental result returned decision tree as the best learner after the performances of K-Nearest Neighbour, Naïve Bayes and Support Vector Machine were tested. The resulting model was built adopting the spiral software engineering framework for the diagnosis system which is deployed on a web based platform.

Cervical Cancer, Machine Learning, Clinical Decision Support System, Diagnosis, Decision Tree

This research considers Polynomial Poisson regression models with orders two and three in one variable for design optimization using constrained design space [0, 1]. The D-optimality criterion is explicitly examined in this study. Imperialist competitive algorithmic procedure was used to generate the optimal design points and weights. The quadratic Poisson regression model was found to be D-optimal at 3-design points: 0.0000, 0.4142 and 1.0000 with design weights 0.3333, 0.3333 and 0.3333 respectively. The cubic Poisson regression model was optimal at 4-design points 0.0000, 0.2204, 0.6596 and 1.0000, each with design weights of 0.25. The constructed D-optimal designs were verified using the general equivalence theorem via the maximum sensitivity function of each model.

D-optimality, Design Point, Fisher Information Matrix and Polynomial Poisson Regression Model

This study attempt an assessment of malaria cases in Nigeria, Malaria is a major public health problem in Nigeria. Most of the malaria infections are dangerous to health, affecting children and adult. Time series analysis was therefore used to formulate the appropriate model for the malaria cases with respect to time and to know whether the trend of malaria is either increasing or decreasing for the period of 1990 to 2020. Trend analysis of the case of malaria reported which indicated that the rate of malaria reported cases is increasing as the year is moving gradually and the reported cases of malaria by 2029 will increase to 443.463 per 1000 population. Malaria is still increasing and causing big problem in the health of children and adult.

Time series, Malaria Infection, Trend Analysis, model, health

The dynamics of research in terms of well-defined study design, interwoven plots/sampling frames, research stages, sample proportions in each plot, nature of output(s) and sample size for different designs can be extensively studied using the theory of mutual information. Owing to the fact that outcomes in multi-dimensional contingency platform are frequentist in nature and as such the usual analysis of variance is not practicable; there is need to develop a sound mathematical techniques for entropy's $H(.)$ significance among interacting vectors whose subsets are discrete and are interwoven into plots in a categorical set up. This study partitions the mutual information algebraic structures via the squared radial function of the exponential power distribution into sum of single and joint entropy's shared contributions in analogy to the analysis of variance sum of squares partitions. We present evidence-based test of entropy significance in few statistical designs with application.

Discrete multi-dimensional design, mutual information, covariance matrix, algebraic structure, entropy significance

One of the most useful brand of mathematics is linear algebra, which have more application in science and engineering, because it required the description of some measurable quantities. In this research paper we determine some application of Eigen-value problems. So in this research work first we discuss how to obtain the Eigen-value and Eigen-vector of square matrix and their characteristic equation and polynomial and then apply the solution of Eigen-value problem to stretching of elastic membrane problems, eigenvalue problems arising from population model and vibrating system of two masses on two springs problems.

Matrix, Eigenvalue, Eigenvector

Residue Number System (RNS) has become one of the most preferred solutions for the implementation of distributed and ubiquitous computing platforms such as cloud, wireless adhoc networks, applications which require tolerance against errors and scalable solutions in mission critical and next generation’s applications. The implementation of RNS for different applications require different approaches depending on the areas of applications. Some applications of Residue Number System described in literature are reviewed so as to illustrate the various· possibilities. Distinct features such as applications, architectures and implementations were considered. In this paper, we presented a comprehensive survey on the different areas of applications, architectures and implementations of RNS to various fields of Information Technology while also including new areas of applications hitherto not covered in previous surveys. The implementations issues with different types of applications such as moduli set, forward conversion, residue arithmetic units, reverse conversion and hardware design, were discussed. Lastly, we focus on the various challenges with the use of RNS and the different solutions that exist and also discuss the future trends in RNS.

Moduli set, forward conversion, reverse conversion, Chinese Remainder Theorem, Mix Radix Conversion