Browsing by Author "Tesha, Thomas"

Now showing 1 - 4 of 4
  • No Thumbnail Available
    Item
    Analysis of Tanzanian biomass consumption using artificial neural network
    (Longdom Publishing SL, 2015) Tesha, Thomas; Kichonge, Baraka
    The growing biomass consumption in developing countries context is being driven by a mixture of concerns over energy security, sustainable development and the climate change mitigation. The development of the comprehensive, sustainable and efficient biomass energy sector policies, strategies and investments requires proper biomass utilization and planning which in fact has not yet received the attention it deserves in the developing countries policies. This paper aims are twofold; one being to demonstrate the practicability of the application of artificial neural network multilayer perceptron (ANN-MLP) in the analysis of the biomass energy consumption and two, to identify the demographic and economic indicators which works better in the analysis and prediction of biomass consumption in Tanzania. Three models made up of Tanzania rural, Tanzania urban and Tanzania population with the addition of economic indicators were formulated for the analysis. The ANN-MLP has shown promising results with the statistical correlation coefficient of 0.9972 indicating that it can be used for practical analysis and prediction of biomass energy consumption. Furthermore the results show the use of Tanzania population model in the analysis and prediction of biomass consumption gives better results in comparison to the Tanzania rural and Tanzania urban population models individually.
  • Thumbnail Image
    Item
    Analysis of Tanzanian energy demand using artificial neural network and multiple linear regression
    (Foundation of Computer Science, 2014) Kichonge, Baraka; Tesha, Thomas; Mkilaha, Iddi S.N.; John, Geoffrey R
    Analysis of energy demand is of a vital concern to energy systems analysts and planners in any nation. This paper present artificial neural network-multilayer perceptron (ANNMLP) and multiple linear regression (MLR) techniques for the analysis of energy demand in Tanzania. The techniques were employed to analyze the influence of economic, energy and environment indicators models in predicting the energy demand in Tanzania. Statistical performance indices were used to evaluate the prediction ability of economic, energy and environment indicators models using ANN-MLP and MLR techniques. Predicted responses values of ANN-MLP and MLR techniques were then compared to determine their closeness with actual data values for determining the best performing technique. The results from ANN-MLP and MLR techniques showed the best model for predicting the energy demand in Tanzania were from energy indicators as opposed to economic and environmental indicators. The ANN-MLP prediction values had a correlation coefficient (CC) of 0.9995 and mean absolute percentage error (MAPE) of 0.67% outperforming the MLR technique whose CC and MAPE values were 0.9993 and 0.83% respectively. ANN-MLP technique graphical presentation of actual against predicted values showed close relationship between actual and predicted values as opposed to the MLR technique whose predicted values deviated much from actual values. Analysis of results from both techniques conclude that ANN-MLP outperform MLR technique in predicting energy demand in Tanzania.
  • Thumbnail Image
    Item
    The impact of transformed features in automating the Swahili document classification
    (Foundation of Computer Science, 2015) Tesha, Thomas
    This paper describes experimental results in an attempt to identify the Transformation techniques which can be adopted to improve features for the automation of classification of Swahili documents. This means improving classification rate by enhancing separability and accuracy. The experiment involved Relative Frequency (RF), Power transformation (PT) and Relative Frequency with Power transformation (RFPT). The Term weighting with TFIDF and the absolute features (AF) were also studied. The features’ dimension reduction was done by using the statistical techniques of Principal Component Analysis. In learning algorithm, the Support vector machine for classification and the k-NN were used, and in evaluating the effect of features’ performance with the classifiers the micro averaged f-measure were adopted. The extensive experimental results demonstrated that the RFPT features worked better with the Support Vector Machine classifiers unlike k-NN in improving the classification rate by enhancing document separability and accuracy in Automation of Swahili document classification.
  • Thumbnail Image
    Item
    Prediction of Tanzanian energy demand using support vector machine for regression (SVR)
    (Taylor and Francis Ltd, 2015) Kichonge, Baraka; John, Geoffrey R.; Tesha, Thomas; Mkilaha, Iddi S.N.
    This study discusses the influences of economic, energy and environment indicators in the prediction of energy demand for Tanzania applying support vector machine for regression (SVR). Economic, energy and environment indicators were applied to formulate models based on time series data. The experimental results showed the supremacy of the polynomial-SVR kernel function and the energy indicators model in providing the transformation, which achieved more accurate prediction values. The energy indicators model had a correlation coefficient (CC) of 0.999 as equated to 0.9975 and 0.9952 with PUKF-SVR kernels for economic and environment indicators model. The energy indicators model closeness of predicted values as compared to actual values was the best as compared to economic and environment indicators models. Furthermore, root mean squared error (RMSE), mean absolute error (MAE), root relative squared error (RRSE) and relative absolute error (RAE) of energy indicators model were the lowest. Long-run sustainable development of the energy sector can be achieved with the use of SVR-algorithm as prediction tool of future energy demand.