Indonesian Journal of Electrical Engineering and Computer Science Vol. 24, No. 2, November 2021, pp. 1009~1016 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v24.i2.pp1009-1016  1009 Journal homepage: http://ijeecs.iaescore.com Dropout, a basic and effective regularization method for a deep learning model: a case study Brahim Jabir, Noureddine Falih LIMATI Laboratory, Sultan Moulay Slimane University, Beni Mellal, Morocco Article Info ABSTRACT Article history: Received May 3, 2021 Revised Sep 8, 2021 Accepted Sep 13, 2021 Deep learning is based on a network of artificial neurons inspired by the human brain. This network is made up of tens or even hundreds of "layers" of neurons. The fields of application of deep learning are indeed multiple; Agriculture is one of those fields in which deep learning is used in various agricultural problems (disease detection, pest detection, and weed identification). A major problem with deep learning is how to create a model that works well, not only on the learning set but also on the validation set. Many approaches used in neural networks are explicitly designed to reduce overfit, possibly at the expense of increasing validation accuracy and training accuracy. In this paper, a basic technique (dropout) is proposed to minimize overfit, we integrated it into a convolutional neural network model to classify weed species and see how it impacts performance, a complementary solution (exponential linear units) are proposed to optimize the obtained results. The results showed that these proposed solutions are practical and highly accurate, enabling us to adopt them in deep learning models. Keywords: CNN Deep learning Dropout Machine learning Regularization This is an open access article under the CC BY-SA license. Corresponding Author: Brahim Jabir LIMATI Laboratory Sultan Moulay Slimane University Mghila, BP 592 Beni Mellal, Morocco Email: ibra.jabir@gmail.com 1. INTRODUCTION Machine learning and deep learning are part of artificial intelligence. These approaches both result in empowering computers to make intelligent decisions. However, deep learning is a subcategory of machine learning because it relies on unattended learning, which is a form of learning based on mathematical approaches used to model data [1]. Deep learning applications are used in various sectors like: image recognition, automatic translation, autonomous car, medical diagnosis, personalized recommendations, automatic moderation of social networks, financial prediction and automated trading, identification of defective parts. The field that interests us and in which we have experimented is “agriculture” [2], [3], indeed, we can use it for the detection of weeds, water management, detection of insects and diseases. Deep learning is a network that is made up of tens or even hundreds of “layers” of neurons, each receiving and interpreting information from the previous layer [4]. A set of theories and models have been brought in existence. Their unified goal is to reach higher accuracy levels that can be applied to solve problems in several fields, agricultural, industrial, health. This field always remains a subject of research. Thousands of experiments and thousands of scientific papers have been produced in deep learning and machine learning in all application fields in recent years. However, the research door is still open where scientists are still trying to reach better models that can gain an important degree of learning. They are trying to discover all the settings that affect it, starting from data collection, through its preparation and purification,