The purpose of this study is to investigate mass valuation of unimproved land value using machine learning techniques. The study was conducted in Nairobi County. It is one of the 47 Kenyan Counties under the 2010 cons...The purpose of this study is to investigate mass valuation of unimproved land value using machine learning techniques. The study was conducted in Nairobi County. It is one of the 47 Kenyan Counties under the 2010 constitution. A total of 1440 geocoded data points containing the market selling price of vacant land in Nairobi were web scraped from major property listing websites. These data points were adopted as dependent variables given as unit price of vacant land per square meter. The Covariates used in this study were categorized into Accessibility, Environmental, Physical and Socio-Economic Factors. Due to multi-collinearity problem present in the covariates, PLS and PCA methods were adopted to transform the observed features using a set of vectors. These methods resulted in an uncorrelated set of components that were used in training machine learning algorithms. The dependent variable and uncorrelated components derived feature reduction methods were used as training data for training different machine learning regression models namely;Random forest, support vector regression and extreme gradient boosting regression (XGboost regression). PLS performed better than PCA because the former maximizes the covariance between dependent and independent variables while the latter maximizes variance between the independent variables only and ignores the relationship between predictors and response. The first 9 components were identified as significant both by PLS and PCA methods. The spatial distribution of vacant land value within Nairobi County was consistent for all the three machine learning models. It was also noted that the land value pattern was higher in the central business district and the pattern spread northwards and westwards relative to the CBD. A relative low vacant land value pattern was observed on the eastern side of the county and also at the extreme periphery of Nairobi County boundary. From the accuracy metrics of R-squared and MAPE, Random Forest Regression model performed better than XGBoost and SVR models. This confirms the capability of random forest model to predict valid estimates of vacant land value for purposes of property taxation in Nairobi County.展开更多
Monitoring Forest degradation is evidence enough to show a country’s commitment to monitor the forest trend both for national and local decision-making and international reporting processes. Unlike deforestation whic...Monitoring Forest degradation is evidence enough to show a country’s commitment to monitor the forest trend both for national and local decision-making and international reporting processes. Unlike deforestation which is easier to point out, monitoring forest degradation is quite a challenge since there is no universal definition and thus no clear monitoring methods apart from the canopy cover change. This research, therefore, sought to look at the degradation trends in the Mau forest complex between 1995-2020 with the aim of finding out whether monitoring canopy density changes over time and quantifying these changes in terms of biomass loss could be a good approach in monitoring forest degradation. Forest Canopy Density (FCD) model was adopted focusing on using vegetation indices describing biophysical conditions of Vegetation, Shadow and Bareness to monitor changes in canopy density as a parameter for describing forest degradation in the forest blocks of Maasai Mau and Olpusimoru in Mau forest complex. Results indicated how different vegetation indices responded to changes in the vegetation density and eventually changes in the canopy density values which were converted in terms of biomass loss. The forest Canopy Density model proved to be a good tool for monitoring forest degradation since it combines different biophysical indices with different characteristics capturing what is happening below the canopy.展开更多
Water scarcity is currently still a global challenge despite the fact that water sustains life on earth. An understanding of domestic water demand is therefore vital for effective water management. In order to underst...Water scarcity is currently still a global challenge despite the fact that water sustains life on earth. An understanding of domestic water demand is therefore vital for effective water management. In order to understand and predict future water demand, appropriate mathematical models are needed. The present work used Geographic Information Systems (GIS) based regression models;Geographically weighted regression (GWR) and Ordinary Least Square (OLS) to model domestic water demand in Athi river town. We identified a total of 7 water determinant factors in our study area. From these factors, 4 most significant ones (household size, household income, meter connections and household rooms) were identified using OLS. Further, GWR technique was used to investigate any intrinsic relationship between the factors and water demand occurrence. GWR coefficients values computed were mapped to exhibit the relationship and strength of each explanatory variable to water demand. By comparing OLS and GWR models with both AIC value and R2 value, the results demonstrated GWR model as capable of projecting water demand compared to OLS model. The GWR model was therefore adopted to predict water demand in the year 2022. It revealed domestic water demand in 2017 was estimated at 721,899 m3 compared to 880,769 m3 in 2022, explaining an increase of about 22%. Generally, the results of this study can be used by water resource planners and managers to effectively manage existing water resources and as baseline information for planning a cost-effective and reliable water supply sources to the residents of a town.展开更多
文摘The purpose of this study is to investigate mass valuation of unimproved land value using machine learning techniques. The study was conducted in Nairobi County. It is one of the 47 Kenyan Counties under the 2010 constitution. A total of 1440 geocoded data points containing the market selling price of vacant land in Nairobi were web scraped from major property listing websites. These data points were adopted as dependent variables given as unit price of vacant land per square meter. The Covariates used in this study were categorized into Accessibility, Environmental, Physical and Socio-Economic Factors. Due to multi-collinearity problem present in the covariates, PLS and PCA methods were adopted to transform the observed features using a set of vectors. These methods resulted in an uncorrelated set of components that were used in training machine learning algorithms. The dependent variable and uncorrelated components derived feature reduction methods were used as training data for training different machine learning regression models namely;Random forest, support vector regression and extreme gradient boosting regression (XGboost regression). PLS performed better than PCA because the former maximizes the covariance between dependent and independent variables while the latter maximizes variance between the independent variables only and ignores the relationship between predictors and response. The first 9 components were identified as significant both by PLS and PCA methods. The spatial distribution of vacant land value within Nairobi County was consistent for all the three machine learning models. It was also noted that the land value pattern was higher in the central business district and the pattern spread northwards and westwards relative to the CBD. A relative low vacant land value pattern was observed on the eastern side of the county and also at the extreme periphery of Nairobi County boundary. From the accuracy metrics of R-squared and MAPE, Random Forest Regression model performed better than XGBoost and SVR models. This confirms the capability of random forest model to predict valid estimates of vacant land value for purposes of property taxation in Nairobi County.
文摘Monitoring Forest degradation is evidence enough to show a country’s commitment to monitor the forest trend both for national and local decision-making and international reporting processes. Unlike deforestation which is easier to point out, monitoring forest degradation is quite a challenge since there is no universal definition and thus no clear monitoring methods apart from the canopy cover change. This research, therefore, sought to look at the degradation trends in the Mau forest complex between 1995-2020 with the aim of finding out whether monitoring canopy density changes over time and quantifying these changes in terms of biomass loss could be a good approach in monitoring forest degradation. Forest Canopy Density (FCD) model was adopted focusing on using vegetation indices describing biophysical conditions of Vegetation, Shadow and Bareness to monitor changes in canopy density as a parameter for describing forest degradation in the forest blocks of Maasai Mau and Olpusimoru in Mau forest complex. Results indicated how different vegetation indices responded to changes in the vegetation density and eventually changes in the canopy density values which were converted in terms of biomass loss. The forest Canopy Density model proved to be a good tool for monitoring forest degradation since it combines different biophysical indices with different characteristics capturing what is happening below the canopy.
文摘Water scarcity is currently still a global challenge despite the fact that water sustains life on earth. An understanding of domestic water demand is therefore vital for effective water management. In order to understand and predict future water demand, appropriate mathematical models are needed. The present work used Geographic Information Systems (GIS) based regression models;Geographically weighted regression (GWR) and Ordinary Least Square (OLS) to model domestic water demand in Athi river town. We identified a total of 7 water determinant factors in our study area. From these factors, 4 most significant ones (household size, household income, meter connections and household rooms) were identified using OLS. Further, GWR technique was used to investigate any intrinsic relationship between the factors and water demand occurrence. GWR coefficients values computed were mapped to exhibit the relationship and strength of each explanatory variable to water demand. By comparing OLS and GWR models with both AIC value and R2 value, the results demonstrated GWR model as capable of projecting water demand compared to OLS model. The GWR model was therefore adopted to predict water demand in the year 2022. It revealed domestic water demand in 2017 was estimated at 721,899 m3 compared to 880,769 m3 in 2022, explaining an increase of about 22%. Generally, the results of this study can be used by water resource planners and managers to effectively manage existing water resources and as baseline information for planning a cost-effective and reliable water supply sources to the residents of a town.
