الفهرس 
Chapter 1: IntroductionOnly 14 pages are availabe for public view
 |  | from | 151 |  |  |
| | | from | 151 | | |
Abstract
Recently, Agriculture is considered among the key strengths of the global country’s economy. The
practice of farming is one of the main occupations in the world and the product key to the variety
of crops. Recently, agriculture is facing problems that may threaten its future, such as drought, and
crop diseases. The world’ s population is growing by about three people per second, equivalent to
250,000 people per day, and by 2025 the world’s population will reach 8 billion, and the plane t’ s
population is expected to reach about 9.6 billion in 2050, according to Food and Agriculture
Organization (FAO) of the United Nations to keep pace with this steady increase, farmers must
upturn food production while conserving the environment and use natural resources rationally, but
they cannot do it on their own, and traditional farming techniques do not enable them to do. So,
modern technologies play an essential role in helping to face the growing food needs of the
world’s population. To do that we suggest a smart farming system with an efficient prediction
method called WPART based on machine learning techniques to predict crop productivity and
drought for professional decision support making innovative farming systems. In this work, an
intelligent manner based on the blend of a wrapper feature selection approach, and PART
classification technique is proposed for crop productivity and drought predicting.
Then we deal with rice diseases that may lead to no grain harvest. This leads researchers to
think about fast, intuitive, most effective, inexpensive, and accurate ways to detect rice diseases
using modern technologies like Artificial Intelligence (AI). Therefore, we propose an effective
rice disease detection and forecasting framework using machine learning and deep learning
techniques in intelligent farming systems for superlative decision-making based on computer
vision and the principle of image processing. Finally, machine learning (ML) requires a large,
comprehensive, and consistent dataset. However, the availability of these datasets is limited because
datasets often suffer from incomplete due to missing data corresponding to different input features,
making developing powerful predictive models based on the ML challenge. Missing data is an
ongoing issue during the development, evaluation, and implementation of predicting models. We are
trying to determine if and to what extent the prediction model studies using the machine learning
report on their treatment of missing data.