Pierre Clement CYEMEZO - Academia.edu (original) (raw)

Uploads

Papers by Pierre Clement CYEMEZO

International Journal of Scientific and Research Publications (IJSRP), 2019

The forecasts are prediction, projection, or estimate of some future activity derived from observ... more The forecasts are prediction, projection, or estimate of some future activity derived from observations that are random variables, and they are not especially informative unless accompanied by some indication of the magnitude of prediction error. Several researchers using different tools such as SPSS, WEKA and Python with various algorithms and models, worked on forecast of agricultural areas include crop yield. However, their findings still need improvement for more accurate performance. The main objective of this study is to forecast crop yield value, analyze predictions and reveal factors that influence the variability of that crop yield value for Rwanda. To achieve this objective, a dataset on crop (Banana, Beans, Cassava, Maize, Rice, Coffee and Tea) yield extracted from Food and Agriculture Organization (FAO) online database on Rwanda between 1990 and 2017 is forecasted using WEKA 3.8.3 Time series forecasting Multilayer Perceptron algorithm with Mean Absolute Percentage Error (MAPE) accuracy metric. From predictions, we found that there is a good crop yield in Rwanda with a growth rate of 38.42%, and we estimate to achieve 121096.66hg/ha by 2030; The best growing yields are Rice with 81.95% and Cassava with 80.9% growth rate in 41 years period and for home consumables while Tea is the best growing exported crop with 34.79% growth rate. By averaging the yield of crops from 1990 to 2030, the highest yield value is for Potatoes with 90794.51hg/ha. Political factors such Genocide against Tutsi in 1994, introduction of technology and fertilizers in agriculture; disasters and climate changes are the main factors that influenced Rwandan crop yield from 1990 to 2030.

International Journal of Scientific and Research Publications (IJSRP), 2019

The forecasts are prediction, projection, or estimate of some future activity derived from observ... more The forecasts are prediction, projection, or estimate of some future activity derived from observations that are random variables, and they are not especially informative unless accompanied by some indication of the magnitude of prediction error. Several researchers using different tools such as SPSS, WEKA and Python with various algorithms and models, worked on forecast of agricultural areas include crop yield. However, their findings still need improvement for more accurate performance. The main objective of this study is to forecast crop yield value, analyze predictions and reveal factors that influence the variability of that crop yield value for Rwanda. To achieve this objective, a dataset on crop (Banana, Beans, Cassava, Maize, Rice, Coffee and Tea) yield extracted from Food and Agriculture Organization (FAO) online database on Rwanda between 1990 and 2017 is forecasted using WEKA 3.8.3 Time series forecasting Multilayer Perceptron algorithm with Mean Absolute Percentage Error (MAPE) accuracy metric. From predictions, we found that there is a good crop yield in Rwanda with a growth rate of 38.42%, and we estimate to achieve 121096.66hg/ha by 2030; The best growing yields are Rice with 81.95% and Cassava with 80.9% growth rate in 41 years period and for home consumables while Tea is the best growing exported crop with 34.79% growth rate. By averaging the yield of crops from 1990 to 2030, the highest yield value is for Potatoes with 90794.51hg/ha. Political factors such Genocide against Tutsi in 1994, introduction of technology and fertilizers in agriculture; disasters and climate changes are the main factors that influenced Rwandan crop yield from 1990 to 2030.