My research areas include Bayesian Analysis, Computational statistics, Artificial Intelligence and Machine Learning and Time series. I am well organized, reliable and willing to work at any time. I am skillful in the use of computer software packages such as SPSS, E-views, R, Stata, python and other statistical packages. Supervisors: Professor Abosede Adedayo Adepoju Phone: 08062483326 Address: Room 3, Department of Statistics New Building, Faculty of Science, University of Ibadan.
Background: Providing shelter and security through housing is a basic need for people. However, t... more Background: Providing shelter and security through housing is a basic need for people. However, the housing market is frequently troubled by issues such as exorbitant rent pricing and a lack of transparency, posing challenges for both landlords and home renters. Objective: This study examines the factors affecting residential housing rent in Ibadan City to thoroughly investigate the cost-effectiveness of rents. We looked into a number of variables, including neighborhood characteristics, property size, amenities, and location, aiming to pinpoint the essential elements significantly impacting rental costs. Methods: Data were collected from two well-known real estate websites, privateproperty.com and propertypro.com. We built a rent prediction model and determined key features using regression analysis utilizing Ordinary Least Squares (OLS), and the Extreme Gradient Boosting (XGBoost) algorithm. The analysis revealed several factors with notable relationships to rent prices. Results: The five variables with the highest correlation coefficients were identified as duplex (0.54), toilet (0.34), detached (0.31), flat (-0.3), and bedroom (0.29), indicating their substantial influence on housing rental rates. Furthermore, the study highlights specific neighborhoods as either relatively expensive or budget-friendly in terms of median rent prices. Notably, Agodi and Kolapo Ishola emerged as the most expensive areas, surpassing the overall median rent by significant margins (454.55% and 450%, respectively). Conversely, Sanyo, Moniya, and Academy recorded the lowest median rent prices, being 50% lower than the overall median. Conclusion: This study provides valuable insights into the intricate factors governing residential housing rent in Ibadan City, offering a foundation for further research and practical applications in the real estate domain.
Background: The changes in the size of the population take a systemic pattern of variation at dif... more Background: The changes in the size of the population take a systemic pattern of variation at different periods of demographic studies. Objective: This study examines the demographic analysis using different regression estimators to determine the effect of live births and factors responsible for maternal mortality in Oyo State. Methods: Analysis of child state of birth such as cephalic presentation, twin, stillbirth, triplet, and prematurity was carried out using discrete regression estimators (negative binomial (NB) regression, zero-inflated regression, poisson regression, and quasi-poisson regression) to determine the factors responsible for live birth and maternal mortality in the state. Level of significance of 5% was used to authenticate the results. Results: The result revealed that cephalic presentation, triplet, and prematurity significantly contributed to live birth at 5%. In contrast, estimated results showed that twin, stillbirth, and triplet states of a child contributed...
This is an open access article under the CC-BY-SA license. Keywords artificial neural network dro... more This is an open access article under the CC-BY-SA license. Keywords artificial neural network drought indices mean square error activation functions maize yield 20
Background: Providing shelter and security through housing is a basic need for people. However, t... more Background: Providing shelter and security through housing is a basic need for people. However, the housing market is frequently troubled by issues such as exorbitant rent pricing and a lack of transparency, posing challenges for both landlords and home renters. Objective: This study examines the factors affecting residential housing rent in Ibadan City to thoroughly investigate the cost-effectiveness of rents. We looked into a number of variables, including neighborhood characteristics, property size, amenities, and location, aiming to pinpoint the essential elements significantly impacting rental costs. Methods: Data were collected from two well-known real estate websites, privateproperty.com and propertypro.com. We built a rent prediction model and determined key features using regression analysis utilizing Ordinary Least Squares (OLS), and the Extreme Gradient Boosting (XGBoost) algorithm. The analysis revealed several factors with notable relationships to rent prices. Results: The five variables with the highest correlation coefficients were identified as duplex (0.54), toilet (0.34), detached (0.31), flat (-0.3), and bedroom (0.29), indicating their substantial influence on housing rental rates. Furthermore, the study highlights specific neighborhoods as either relatively expensive or budget-friendly in terms of median rent prices. Notably, Agodi and Kolapo Ishola emerged as the most expensive areas, surpassing the overall median rent by significant margins (454.55% and 450%, respectively). Conversely, Sanyo, Moniya, and Academy recorded the lowest median rent prices, being 50% lower than the overall median. Conclusion: This study provides valuable insights into the intricate factors governing residential housing rent in Ibadan City, offering a foundation for further research and practical applications in the real estate domain.
In a study to compare the durability of commonly used stains (Giemsa, Leishman, Wright, Eosin, Ni... more In a study to compare the durability of commonly used stains (Giemsa, Leishman, Wright, Eosin, Nigrosin and Gentian violet) for exfoliative vaginal cytology, vaginal smear was obtained from eleven apparently healthy West African Dwarf (WAD) female Goats and processed according to standard technique. Scores (0-3) were given on four parameters namely background of smears, overall staining pattern, cytoplasmic staining and nuclear staining. Quality index one (QI-I) was calculated from the ratio of score achieved to the maximum score possible (12), immediately post staining while quality index–II (QI-II) was obtained 35 days after. Calculation for durability index (DI) was self-derived and equalled to ratio of QI-II to QI-I. The data were presented as mean ± SEM. Multinomial logistic regression model was generated for the QI-I and QI-II using durability index as reference category. Giemsa, Leishman and Wright stains were more durable than others with their mean DI values significantly (...
Background: A cost reflective tariff eliminates the need for government subsidies to make up the ... more Background: A cost reflective tariff eliminates the need for government subsidies to make up the difference between the prevailing tariff and the actual cost of supply by reflecting the true cost of producing power. However, in order to provide adequate cost reflective tariff, there must be a balanced understanding between the electricity producers and the users in terms of electricity production and supply. It is against this backdrop that this study seeks to investigate the relationship between cost reflective electricity tariff and distribution in Nigeria. Objective: The aim of this study is to empirically investigate the impact of cost reflective tariff on the Nigerian electricity supply industry and to investigate the relationship between cost reflective tariff and electricity distribution using multiple regression analyses. Methods: The data used in this study were secondary data generated and extracted from the World Bank indicators (2020), CBN statistical bulletin (2020) and supplemented with data from Nigerian Electricity Regulatory Commission (2020) that covered the period of the study for the 6 states in southwest. Two models were specified to address the two specific research objectives. The first model sought to test the impact of cost reflective tariff on the Nigeria electricity supply industry using multiple regression analyses and the other to investigate the factors that contributed to the price of electricity determinants in Nigeria. Results: The results showed that the cost of electricity production contributed significantly to the quantity of electricity distributed and the model produced R-square and adjusted R-square values of 91.16% and 87.38%, respectively. Conclusion: The results obtained in this work is useful for electricity producers and government for proper planning.
In a study to compare the durability of commonly used stains (Giemsa, Leishman, Wright, Eosin, Ni... more In a study to compare the durability of commonly used stains (Giemsa, Leishman, Wright, Eosin, Nigrosin and Gentian violet) for exfoliative vaginal cytology, vaginal smear was obtained from eleven apparently healthy West African Dwarf (WAD) female Goats and processed according to standard technique. Scores (0-3) were given on four parameters namely background of smears, overall staining pattern, cytoplasmic staining and nuclear staining. Quality index one (QI-I) was calculated from the ratio of score achieved to the maximum score possible (12), immediately post staining while quality index-II (QI-II) was obtained 35 days after. Calculation for durability index (DI) was self-derived and equalled to ratio of QI-II to QI-I. The data were presented as mean ± SEM. Multinomial logistic regression model was generated for the QI-I and QI-II using durability index as reference category. Giemsa, Leishman and Wright stains were more durable than others with their mean DI values significantly (P < 0.05) higher than Gentian violet, Nigrosin and Eosin.The model showed 89.2 % overall model accuracy for the multinomial logistic regression model and 81.5% for the multinomial Bayes Naïve regression model. In conclusion, Giemsa, Leishman and Wright stains were more reliable and durable for exfoliative vaginal cytology compared to the other stains.
The existence of outliers in the original sample may create problem to the classical bootstrappin... more The existence of outliers in the original sample may create problem to the classical bootstrapping estimates. There is possibility that the bootstrap samples may contain more outliers than the original dataset since the bootstrap re-sampling is with replacement. Consequently, the outliers will have an unduly effect on the classical bootstrap mean and standard deviation. This study examined the performance of three bootstrap estimators namely: Case Bootstrapping, Fixed-X Bootstrapping and Residual Resampling method under different levels of outliers. The objective was to determine which of these bootstrap methods is resistant to the presence of outliers in the data. Three levels of outliers; 5%, 10% and 20% were considered and injected into sample sizes, N = 20, 30, 50, and 100 each replicated 1000 and 5000 times respectively. The performances of the bootstrap methods were evaluated using the mean, standard error, absolute bias, mean square error and the root mean square error. The results showed that the Residual resampling Bootstrap performed better than the other two estimators.
Terrorism continues to be one of the most important threats to today's civilization. The differen... more Terrorism continues to be one of the most important threats to today's civilization. The different forms of terrorist attacks in Nigeria in the recent times are Boko-Haram attack, Fulani/Herdsmen attack, Inter/Intra-group conflicts, robbery and lack of intentionality. In order to curb or reduce these activities in Nigeria, there is a need to develop models that can be used to understand these terrorists' activities and prevent or reduce future occurrences. Objective: The aim of this work is to predict terrorist activities in Nigeria using machine learning models (MLM). Methods: The data used in this study was gathered from the daily terrorism incidents throughout Nigeria. The data consist of the different kinds of attacks, the success and the suicidal rates of the attacks and the different levels of weapon types used during the attacks. The targets or victims of the terrorist attack, perpetrators information, casualties and the incidents' consequences were also the highlights in the database. A Heterogeneous Neural Network(HETNN) model was used and its performance was compared with five other MLM namely: Logistic regression (LR), Support Vector Machine(SVM), K-Nearest Neighbour (KNN), Boosting and Random Forest classification models. Findings: The results show that HETNN performs better in prediction compared to the other models. It was also discovered that in determining the success of a terrorist attack, the factors to be considered in order of importance are the number of perpetrators, attack type, type of weapon, the type of victims targeted, and the state of the incidents. Conclusion: The information provided in this work will help the Nigeria government and the security agents in combating insecurity issues in the country.
Neural networks have been very important models across computer
vision, natural language processi... more Neural networks have been very important models across computer vision, natural language processing, speech and image recognition, aircraft safety and many more. It uses a variety of architectures that centres on the Multi-Layer Perceptron (MLP) which is the most commonly used type of Artificial Neural Network. MLP has been found to be good in terms of model precision in the usage of Homogenous Transfer/activation Functions (HTFs), especially with large data set. Based on the preliminary investigations of ranking of transfer functions by error variance (Udomboso, 2014), three HTFs are considered to perform better than other HTFs in prediction. These HTFs are the Hyperbolic Tangent Transfer functions (TANH), Hyperbolic Tangent Sigmoid Transfer function (TANSIG) and the Symmetric Saturating Linear Transfer Function (SSLTF). In this work, the performance of two Heterogeneous Transfer Functions (HETFs), which came as a result of the convolution of the three best HTFs, were compared with the performance of the three above listed HTFs. The hidden neurons used are 2, 5 and 10, while the sample sizes include 50, 100, 200, 500 and 1000. The data were divided into training sets of 90, 80 and 70 respectively. The results showed that the HETFs performed better in terms of the forecast using Mean Square Error (MSE), Mean Absolute Error (MAE) and Test Error as the forecast prediction criteria.
It has been observed over the years that real life data are usually non-conforming to the classic... more It has been observed over the years that real life data are usually non-conforming to the classical linear regression assumptions. One of the stringent assumptions that is unlikely to hold in many applied settings is that of homoscedasticity. When homogenous variance in a normal regression model is not appropriate, invalid standard inference procedure may result from the improper estimation of standard error when the disturbance process in a regression model present heteroscedasticity. When both outliers and heteroscedasticity exist, the inflation of the scale estimate can deteriorate. This study identifies outliers under heteroscedastic errors and seeks to study the performance of four methods; ordinary least squares (OLS), weighted least squares (WLS), robust weighted least squares (RWLS) and logarithmic transformation (Log Transform) methods to estimate the parameters of the regression model in the presence of heteroscedasticity and outliers. Real life data obtained from the Central Bank of Nigeria Bulletin and Monte Carlo simulation were carried out to investigate the performances of these four estimators. The results obtained show that the transformed logarithmic model proved to be the best estimator with minimum standard error followed by the robust weighted least squares. The performance of OLS is the least in this order.
Research & Reviews: Journal of Statistics and Mathematical Sciences, 2022
Drought is a key abiotic stress affecting maize yield and production in Sub Saharan Africa contri... more Drought is a key abiotic stress affecting maize yield and production in Sub Saharan Africa contributing between 44% to 58% grain yield decline in West and Central Africa. For the detection, classification, and control of drought conditions, drought indices are used. This paper presents the application of a multiple linear regression model and spatial distribution to assess the performance of drought indices on maize production in the Northern part of Nigeria. In this research, observed annual data of drought indices, RDI and the palmer drought indices which includes SCPDSI, SCPHDI and SCWLPM, maize yield (measured in tonnes) in Northern states of Nigeria were obtained from 1993 to 2018. The multiple linear regression was carried out at different training sets: 70%, 80% and 90%. Results from the multiple linear regression showed that in the North-Central states, FCT has the lowest MSE (0.7788234) at 90% training level. In NorthEastern states, Borno state has the lowest MSE (0.7240276) at 80% training sets. In NorthWestern states, Kebbi state has the lowest MSE (0.8029484) at 70% training set. Results from the spatial distribution revealed that Yobe state has the lowest maize yield in the Northern states.
Background: Providing shelter and security through housing is a basic need for people. However, t... more Background: Providing shelter and security through housing is a basic need for people. However, the housing market is frequently troubled by issues such as exorbitant rent pricing and a lack of transparency, posing challenges for both landlords and home renters. Objective: This study examines the factors affecting residential housing rent in Ibadan City to thoroughly investigate the cost-effectiveness of rents. We looked into a number of variables, including neighborhood characteristics, property size, amenities, and location, aiming to pinpoint the essential elements significantly impacting rental costs. Methods: Data were collected from two well-known real estate websites, privateproperty.com and propertypro.com. We built a rent prediction model and determined key features using regression analysis utilizing Ordinary Least Squares (OLS), and the Extreme Gradient Boosting (XGBoost) algorithm. The analysis revealed several factors with notable relationships to rent prices. Results: The five variables with the highest correlation coefficients were identified as duplex (0.54), toilet (0.34), detached (0.31), flat (-0.3), and bedroom (0.29), indicating their substantial influence on housing rental rates. Furthermore, the study highlights specific neighborhoods as either relatively expensive or budget-friendly in terms of median rent prices. Notably, Agodi and Kolapo Ishola emerged as the most expensive areas, surpassing the overall median rent by significant margins (454.55% and 450%, respectively). Conversely, Sanyo, Moniya, and Academy recorded the lowest median rent prices, being 50% lower than the overall median. Conclusion: This study provides valuable insights into the intricate factors governing residential housing rent in Ibadan City, offering a foundation for further research and practical applications in the real estate domain.
Background: The changes in the size of the population take a systemic pattern of variation at dif... more Background: The changes in the size of the population take a systemic pattern of variation at different periods of demographic studies. Objective: This study examines the demographic analysis using different regression estimators to determine the effect of live births and factors responsible for maternal mortality in Oyo State. Methods: Analysis of child state of birth such as cephalic presentation, twin, stillbirth, triplet, and prematurity was carried out using discrete regression estimators (negative binomial (NB) regression, zero-inflated regression, poisson regression, and quasi-poisson regression) to determine the factors responsible for live birth and maternal mortality in the state. Level of significance of 5% was used to authenticate the results. Results: The result revealed that cephalic presentation, triplet, and prematurity significantly contributed to live birth at 5%. In contrast, estimated results showed that twin, stillbirth, and triplet states of a child contributed...
This is an open access article under the CC-BY-SA license. Keywords artificial neural network dro... more This is an open access article under the CC-BY-SA license. Keywords artificial neural network drought indices mean square error activation functions maize yield 20
Background: Providing shelter and security through housing is a basic need for people. However, t... more Background: Providing shelter and security through housing is a basic need for people. However, the housing market is frequently troubled by issues such as exorbitant rent pricing and a lack of transparency, posing challenges for both landlords and home renters. Objective: This study examines the factors affecting residential housing rent in Ibadan City to thoroughly investigate the cost-effectiveness of rents. We looked into a number of variables, including neighborhood characteristics, property size, amenities, and location, aiming to pinpoint the essential elements significantly impacting rental costs. Methods: Data were collected from two well-known real estate websites, privateproperty.com and propertypro.com. We built a rent prediction model and determined key features using regression analysis utilizing Ordinary Least Squares (OLS), and the Extreme Gradient Boosting (XGBoost) algorithm. The analysis revealed several factors with notable relationships to rent prices. Results: The five variables with the highest correlation coefficients were identified as duplex (0.54), toilet (0.34), detached (0.31), flat (-0.3), and bedroom (0.29), indicating their substantial influence on housing rental rates. Furthermore, the study highlights specific neighborhoods as either relatively expensive or budget-friendly in terms of median rent prices. Notably, Agodi and Kolapo Ishola emerged as the most expensive areas, surpassing the overall median rent by significant margins (454.55% and 450%, respectively). Conversely, Sanyo, Moniya, and Academy recorded the lowest median rent prices, being 50% lower than the overall median. Conclusion: This study provides valuable insights into the intricate factors governing residential housing rent in Ibadan City, offering a foundation for further research and practical applications in the real estate domain.
In a study to compare the durability of commonly used stains (Giemsa, Leishman, Wright, Eosin, Ni... more In a study to compare the durability of commonly used stains (Giemsa, Leishman, Wright, Eosin, Nigrosin and Gentian violet) for exfoliative vaginal cytology, vaginal smear was obtained from eleven apparently healthy West African Dwarf (WAD) female Goats and processed according to standard technique. Scores (0-3) were given on four parameters namely background of smears, overall staining pattern, cytoplasmic staining and nuclear staining. Quality index one (QI-I) was calculated from the ratio of score achieved to the maximum score possible (12), immediately post staining while quality index–II (QI-II) was obtained 35 days after. Calculation for durability index (DI) was self-derived and equalled to ratio of QI-II to QI-I. The data were presented as mean ± SEM. Multinomial logistic regression model was generated for the QI-I and QI-II using durability index as reference category. Giemsa, Leishman and Wright stains were more durable than others with their mean DI values significantly (...
Background: A cost reflective tariff eliminates the need for government subsidies to make up the ... more Background: A cost reflective tariff eliminates the need for government subsidies to make up the difference between the prevailing tariff and the actual cost of supply by reflecting the true cost of producing power. However, in order to provide adequate cost reflective tariff, there must be a balanced understanding between the electricity producers and the users in terms of electricity production and supply. It is against this backdrop that this study seeks to investigate the relationship between cost reflective electricity tariff and distribution in Nigeria. Objective: The aim of this study is to empirically investigate the impact of cost reflective tariff on the Nigerian electricity supply industry and to investigate the relationship between cost reflective tariff and electricity distribution using multiple regression analyses. Methods: The data used in this study were secondary data generated and extracted from the World Bank indicators (2020), CBN statistical bulletin (2020) and supplemented with data from Nigerian Electricity Regulatory Commission (2020) that covered the period of the study for the 6 states in southwest. Two models were specified to address the two specific research objectives. The first model sought to test the impact of cost reflective tariff on the Nigeria electricity supply industry using multiple regression analyses and the other to investigate the factors that contributed to the price of electricity determinants in Nigeria. Results: The results showed that the cost of electricity production contributed significantly to the quantity of electricity distributed and the model produced R-square and adjusted R-square values of 91.16% and 87.38%, respectively. Conclusion: The results obtained in this work is useful for electricity producers and government for proper planning.
In a study to compare the durability of commonly used stains (Giemsa, Leishman, Wright, Eosin, Ni... more In a study to compare the durability of commonly used stains (Giemsa, Leishman, Wright, Eosin, Nigrosin and Gentian violet) for exfoliative vaginal cytology, vaginal smear was obtained from eleven apparently healthy West African Dwarf (WAD) female Goats and processed according to standard technique. Scores (0-3) were given on four parameters namely background of smears, overall staining pattern, cytoplasmic staining and nuclear staining. Quality index one (QI-I) was calculated from the ratio of score achieved to the maximum score possible (12), immediately post staining while quality index-II (QI-II) was obtained 35 days after. Calculation for durability index (DI) was self-derived and equalled to ratio of QI-II to QI-I. The data were presented as mean ± SEM. Multinomial logistic regression model was generated for the QI-I and QI-II using durability index as reference category. Giemsa, Leishman and Wright stains were more durable than others with their mean DI values significantly (P < 0.05) higher than Gentian violet, Nigrosin and Eosin.The model showed 89.2 % overall model accuracy for the multinomial logistic regression model and 81.5% for the multinomial Bayes Naïve regression model. In conclusion, Giemsa, Leishman and Wright stains were more reliable and durable for exfoliative vaginal cytology compared to the other stains.
The existence of outliers in the original sample may create problem to the classical bootstrappin... more The existence of outliers in the original sample may create problem to the classical bootstrapping estimates. There is possibility that the bootstrap samples may contain more outliers than the original dataset since the bootstrap re-sampling is with replacement. Consequently, the outliers will have an unduly effect on the classical bootstrap mean and standard deviation. This study examined the performance of three bootstrap estimators namely: Case Bootstrapping, Fixed-X Bootstrapping and Residual Resampling method under different levels of outliers. The objective was to determine which of these bootstrap methods is resistant to the presence of outliers in the data. Three levels of outliers; 5%, 10% and 20% were considered and injected into sample sizes, N = 20, 30, 50, and 100 each replicated 1000 and 5000 times respectively. The performances of the bootstrap methods were evaluated using the mean, standard error, absolute bias, mean square error and the root mean square error. The results showed that the Residual resampling Bootstrap performed better than the other two estimators.
Terrorism continues to be one of the most important threats to today's civilization. The differen... more Terrorism continues to be one of the most important threats to today's civilization. The different forms of terrorist attacks in Nigeria in the recent times are Boko-Haram attack, Fulani/Herdsmen attack, Inter/Intra-group conflicts, robbery and lack of intentionality. In order to curb or reduce these activities in Nigeria, there is a need to develop models that can be used to understand these terrorists' activities and prevent or reduce future occurrences. Objective: The aim of this work is to predict terrorist activities in Nigeria using machine learning models (MLM). Methods: The data used in this study was gathered from the daily terrorism incidents throughout Nigeria. The data consist of the different kinds of attacks, the success and the suicidal rates of the attacks and the different levels of weapon types used during the attacks. The targets or victims of the terrorist attack, perpetrators information, casualties and the incidents' consequences were also the highlights in the database. A Heterogeneous Neural Network(HETNN) model was used and its performance was compared with five other MLM namely: Logistic regression (LR), Support Vector Machine(SVM), K-Nearest Neighbour (KNN), Boosting and Random Forest classification models. Findings: The results show that HETNN performs better in prediction compared to the other models. It was also discovered that in determining the success of a terrorist attack, the factors to be considered in order of importance are the number of perpetrators, attack type, type of weapon, the type of victims targeted, and the state of the incidents. Conclusion: The information provided in this work will help the Nigeria government and the security agents in combating insecurity issues in the country.
Neural networks have been very important models across computer
vision, natural language processi... more Neural networks have been very important models across computer vision, natural language processing, speech and image recognition, aircraft safety and many more. It uses a variety of architectures that centres on the Multi-Layer Perceptron (MLP) which is the most commonly used type of Artificial Neural Network. MLP has been found to be good in terms of model precision in the usage of Homogenous Transfer/activation Functions (HTFs), especially with large data set. Based on the preliminary investigations of ranking of transfer functions by error variance (Udomboso, 2014), three HTFs are considered to perform better than other HTFs in prediction. These HTFs are the Hyperbolic Tangent Transfer functions (TANH), Hyperbolic Tangent Sigmoid Transfer function (TANSIG) and the Symmetric Saturating Linear Transfer Function (SSLTF). In this work, the performance of two Heterogeneous Transfer Functions (HETFs), which came as a result of the convolution of the three best HTFs, were compared with the performance of the three above listed HTFs. The hidden neurons used are 2, 5 and 10, while the sample sizes include 50, 100, 200, 500 and 1000. The data were divided into training sets of 90, 80 and 70 respectively. The results showed that the HETFs performed better in terms of the forecast using Mean Square Error (MSE), Mean Absolute Error (MAE) and Test Error as the forecast prediction criteria.
It has been observed over the years that real life data are usually non-conforming to the classic... more It has been observed over the years that real life data are usually non-conforming to the classical linear regression assumptions. One of the stringent assumptions that is unlikely to hold in many applied settings is that of homoscedasticity. When homogenous variance in a normal regression model is not appropriate, invalid standard inference procedure may result from the improper estimation of standard error when the disturbance process in a regression model present heteroscedasticity. When both outliers and heteroscedasticity exist, the inflation of the scale estimate can deteriorate. This study identifies outliers under heteroscedastic errors and seeks to study the performance of four methods; ordinary least squares (OLS), weighted least squares (WLS), robust weighted least squares (RWLS) and logarithmic transformation (Log Transform) methods to estimate the parameters of the regression model in the presence of heteroscedasticity and outliers. Real life data obtained from the Central Bank of Nigeria Bulletin and Monte Carlo simulation were carried out to investigate the performances of these four estimators. The results obtained show that the transformed logarithmic model proved to be the best estimator with minimum standard error followed by the robust weighted least squares. The performance of OLS is the least in this order.
Research & Reviews: Journal of Statistics and Mathematical Sciences, 2022
Drought is a key abiotic stress affecting maize yield and production in Sub Saharan Africa contri... more Drought is a key abiotic stress affecting maize yield and production in Sub Saharan Africa contributing between 44% to 58% grain yield decline in West and Central Africa. For the detection, classification, and control of drought conditions, drought indices are used. This paper presents the application of a multiple linear regression model and spatial distribution to assess the performance of drought indices on maize production in the Northern part of Nigeria. In this research, observed annual data of drought indices, RDI and the palmer drought indices which includes SCPDSI, SCPHDI and SCWLPM, maize yield (measured in tonnes) in Northern states of Nigeria were obtained from 1993 to 2018. The multiple linear regression was carried out at different training sets: 70%, 80% and 90%. Results from the multiple linear regression showed that in the North-Central states, FCT has the lowest MSE (0.7788234) at 90% training level. In NorthEastern states, Borno state has the lowest MSE (0.7240276) at 80% training sets. In NorthWestern states, Kebbi state has the lowest MSE (0.8029484) at 70% training set. Results from the spatial distribution revealed that Yobe state has the lowest maize yield in the Northern states.
Uploads
Papers by Tayo P . Ogundunmade
vision, natural language processing, speech and image recognition,
aircraft safety and many more. It uses a variety of architectures that
centres on the Multi-Layer Perceptron (MLP) which is the most
commonly used type of Artificial Neural Network. MLP has been found
to be good in terms of model precision in the usage of Homogenous
Transfer/activation Functions (HTFs), especially with large data set.
Based on the preliminary investigations of ranking of transfer functions
by error variance (Udomboso, 2014), three HTFs are considered to
perform better than other HTFs in prediction. These HTFs are the
Hyperbolic Tangent Transfer functions (TANH), Hyperbolic Tangent
Sigmoid Transfer function (TANSIG) and the Symmetric Saturating
Linear Transfer Function (SSLTF). In this work, the performance of two
Heterogeneous Transfer Functions (HETFs), which came as a result of
the convolution of the three best HTFs, were compared with the
performance of the three above listed HTFs. The hidden neurons used
are 2, 5 and 10, while the sample sizes include 50, 100, 200, 500 and
1000. The data were divided into training sets of 90, 80 and 70
respectively. The results showed that the HETFs performed better in
terms of the forecast using Mean Square Error (MSE), Mean Absolute
Error (MAE) and Test Error as the forecast prediction criteria.
vision, natural language processing, speech and image recognition,
aircraft safety and many more. It uses a variety of architectures that
centres on the Multi-Layer Perceptron (MLP) which is the most
commonly used type of Artificial Neural Network. MLP has been found
to be good in terms of model precision in the usage of Homogenous
Transfer/activation Functions (HTFs), especially with large data set.
Based on the preliminary investigations of ranking of transfer functions
by error variance (Udomboso, 2014), three HTFs are considered to
perform better than other HTFs in prediction. These HTFs are the
Hyperbolic Tangent Transfer functions (TANH), Hyperbolic Tangent
Sigmoid Transfer function (TANSIG) and the Symmetric Saturating
Linear Transfer Function (SSLTF). In this work, the performance of two
Heterogeneous Transfer Functions (HETFs), which came as a result of
the convolution of the three best HTFs, were compared with the
performance of the three above listed HTFs. The hidden neurons used
are 2, 5 and 10, while the sample sizes include 50, 100, 200, 500 and
1000. The data were divided into training sets of 90, 80 and 70
respectively. The results showed that the HETFs performed better in
terms of the forecast using Mean Square Error (MSE), Mean Absolute
Error (MAE) and Test Error as the forecast prediction criteria.