الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 119 |  |  |
| | | from | 119 | | |
Abstract
The need for fresh water utilized in agriculture worldwide rises together with the world population growth and the corresponding rise in food demand. At the same time, the Earth is exposed to many droughts as a result of global warming. As a result, it is urgent that existing farming methods be improved upon. It will assist in addressing world crises like climate change and drought as well as the sustainable future of humankind and all living creatures. Better technology will increase yield, hence reducing the likelihood of scarcity of food and malnutrition. This thesis aims mainly to develop a smart irrigation system that utilizes machine learning (ML) algorithms to enhance the productivity and sustainability of agricultural sites by providing the appropriate amount of irrigation water at the optimal time. This thesis presents two smart irrigation models based on machine learning algorithms for detecting the appropriate amount of water required for farming crops and reducing the harmful environmental effects of over-irrigation. The first proposed approach (Model I) is based on monitoring the soil moisture and soil temperature using data collected by satellites and weather forecasts to determine the soil needs water and make the effective decision to operate the irrigation sprinklers or not. ML algorithm of (Model I) combines Support Vector Regression (SVR), Decision Tree Regressor (DTR), Random Forest Regression (RFR), Multilinear Regression (MLR), Ridge regression (RR), and k-means clustering. While the second approach (Model II) is based on classifying the amount of water emitted by the sprinklers, of the proposed intelligent irrigation system, into five levels; Max, High, Medium, Low, and Stop. Four ML algorithms are applied in (Model II) to detect the required amount of water and decide iv the level of operating the irrigation sprinklers. These ML algorithms are support vector machine (SVM), decision tree (DT), SVM with AdaBoost, and DT with AdaBoost.