| Titre : | Development of Synthetic Agricultural Datasets Using Generative AI |
| Auteurs : | Loudjeyna, Auteur ; Khaled Rezeg, Auteur |
| Editeur : | Biskra [Algérie] : Faculté des Sciences Exactes et des Sciences de la Nature et de la Vie, Université Mohamed Khider, 2026 |
| Format : | 1 vol. (281 p.) / couv. ill. en coul / 30 |
| Langues: | Français |
| Langues originales: | Anglais |
| Résumé : |
SynthAI is a project that focuses on the development of an intelligent platform for synthetic data generation and agricultural decision support. Despite the rapid advancement of artificial intelligence technologies, many agricultural and machine learning applications still face challenges related to data availability, quality, privacy, and accessibility. Existing solutions often rely heavily on real-world datasets, which may be limited, sensitive, or difficult to share. In addition, some approaches fail to preserve the statistical properties and realistic relationships present in the original data. To address these challenges, we designed and developed a web-based platform that integrates the CTGAN (Conditional Tabular Generative Adversarial Network) model for generating high-quality synthetic agricultural datasets. The platform provides functionalities such as synthetic data generation, data augmentation, model training and management, dataset evaluation, and crop recommendation support. It also offers statistical and visual validation tools to compare real and synthetic datasets and assess their quality and reliability. This report presents the motivations behind the project, its design and implementation process, and the results achieved, with the objective of improving synthetic data generation and supporting data-driven decision-making in smart agriculture. Keywords: Synthetic Data Generation, CTGAN, Smart Agriculture, Machine Learning, Data Augmentation, Crop Recommendation, Data Validation. |
| Sommaire : |
List of Figures viii Nomenclature 1 Abstract 1 General Introduction 3 1 State Of Art 5 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.2 Traditional Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.2.1 Definition of traditional agriculture . . . . . . . . . . . . . . . . . . . 5 1.2.2 Characteristics of Traditional agriculture . . . . . . . . . . . . . . . . 6 1.2.3 Limitations of Traditional Agriculture . . . . . . . . . . . . . . . . . 6 1.3 Smart Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.3.1 Definition and principles of smart agriculture . . . . . . . . . . . . . 6 1.3.1.1 Definition of smart agriculture . . . . . . . . . . . . . . . . 7 1.3.1.2 Difference Between Traditional and Smart Agriculture . . . 7 1.3.1.3 Principles of Smart Agriculture . . . . . . . . . . . . . . . . 7 1.3.2 Core technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.3.3 Evolution of smart agriculture . . . . . . . . . . . . . . . . . . . . . . 8 1.3.4 Benefits of Smart Agriculture . . . . . . . . . . . . . . . . . . . . . . 9 1.4 Agricultural Practices in Smart Farming ............................................................. 10 1.4.1 Agricultural cycle ...................................................................................... 10 1.4.1.1 Definition .................................................................................. 10 1.4.1.2 Agricultural Cycle in Smart Farming ......................................... 11 1.4.2 Crop Rotation ........................................................................................... 13 1.4.3 Advantages and disadvantages ................................................................. 13 1.5 Fertilization in Modern Agriculture ...................................................................... 14 1.5.1 Definition of fertilizers and fertilization .................................................... 14 1.5.2 Types of fertilizers (Organic, Mineral, NPK, Micronutrients) ................. 14 1.5.3 Essential nutrients and their roles (N, P, K, etc.) ..................................... 15 1.5.4 Challenges in fertilizer management ......................................................... 15 1.5.5 Towards Smart Fertilization ..................................................................... 16 1.6 Fertigation Systems .............................................................................................. 16 1.6.1 Definition and Concept of fertigation ....................................................... 16 1.6.2 Components and Types of fertigation systems ......................................... 17 1.6.3 Advantages and Limitations ..................................................................... 17 1.6.4 Process of Fertilizer Application ............................................................... 18 1.6.5 Smart / Data-Driven Fertigation ............................................................. 18 1.7 Conclusion ....................................................................................... 19 2 Data-Driven Smart Agriculture and Synthetic Data Generation 20 2.1 Introduction ......................................................................................................... 20 2.2 Agricultural Data Infrastructure .......................................................................... 20 2.2.1 Definition of agricultural data ................................................................... 20 2.2.2 Types of agriculture data .......................................................................... 21 2.2.3 Data sources in agriculture ....................................................................... 21 2.3 Machine Learning in Agriculture .......................................................................... 22 2.3.1 Machine learning fundamentals: supervised and unsupervised learning ............................................................................... 22 2.3.2 Agricultural Application: Yield prediction, Disease............................................... 23 2.3.3 Data challenges: Limited datasets, noisy sensors and missing values . 23 2.4 Synthetic Data Generation ................................................................................... 24 2.4.1 Definition of synthetic data ....................................................................... 24 2.4.2 Motivation for synthetic data .................................................................... 24 2.4.3 Evaluation metrics for synthetic data ....................................................... 24 2.5 Generative Models ................................................................................................ 25 2.5.1 Overview of generative models: GAN, VAE and Diffusion models .......... 25 2.5.2 Generative Adversarial Networks (GAN): Generator, Discriminator and Training ............................................................................................. 26 2.5.3 Variational Autoencoders (VAE): Encoder, Latent Space and Decoder 26 2.5.4 CTGAN for Tabular Data: Tabular data generation, Categorical handling and Advantages.............................................................. 26 2.5.5 Large Language Models (LLMs): Natural Language Processing and Decision Support in Agriculture ............................................................... 26 2.6 Synthetic Data For Smart Agriculture ................................................................. 27 2.6.1 Crop Yield Prediction ............................................................................... 27 2.6.2 Climate Simulation ................................................................................... 27 2.6.3 Irrigation Optimization ............................................................................ 27 2.7 Synthetic Data Applications in Fertigation Systems ............................................ 28 2.7.1 Irrigation Scheduling ................................................................................ 28 2.7.2 Fertilizer Dosage Optimization ................................................................. 28 2.7.3 Nutrient Solution Modelling ..................................................................... 29 2.7.4 Soil Moisture Prediction ........................................................................... 29 2.8 Conclusion ................................................................ 29 3 System Analysis and Design 30 3.1 Introduction ........................................................................................... 30 3.2 Requirement Analysis ........................................................................................... 30 3.2.1 Functional Requirements ......................................................................... 30 3.2.2 Non-Functional Requirements ................................................................... 32 3.3 System Modeling .................................................................................................. 34 3.3.1 Actors Identification ................................................................................. 34 3.3.2 Use Case Diagram .................................................................................... 34 3.3.2.1 Farmer Use Case Diagram ......................................................... 34 3.3.2.2 Company Use Case Diagram ..................................................... 35 3.3.2.3 Admin Use Case Diagram ......................................................... 36 3.3.3 Sequence Diagram .................................................................................... 37 3.3.3.1 Admin Model Training Sequence Diagram ................................ 37 3.3.4 Activity Diagram ...................................................................................... 38 3.4 System Architecture ............................................................................................. 41 3.4.1 Global Architecture .................................................................................. 41 3.4.2 CTGAN Framework Design ..................................................................... 42 3.4.3 Why CTGAN Instead of TVAE? ............................................................... 43 3.5 Database Conception............................................................................................ 43 3.5.1 Conceptual Data Model (MCD) ................................................................ 44 3.5.2 Logical Data Model (MLD) ...................................................................... 44 3.6 Conclusion ..................................................................................... 45 4 System Implementation 46 4.1 Introduction ....................................................................................... 46 4.2 Data Exploration and Preprocessing ................................................................... 46 4.2.1 Data Source and Description.................................................................... 46 4.2.2 Future Data Collection in Algeria ............................................................. 47 4.2.3 Dataset Features....................................................................................... 48 4.2.4 Future Local Data Collection ................................................................... 49 4.2.5 Exploratory Data Analysis (EDA) ............................................................ 50 4.2.5.1 Feature Distribution .................................................................. 50 4.2.5.2 Correlation Analysis .................................................................. 51 4.2.5.3 Outlier Detection ....................................................................... 51 4.2.5.4 Crop Class Distribution ............................................................. 52 4.2.5.5 Feature Relationships ................................................................ 53 4.2.6 Data Cleaning ........................................................................................... 53 4.2.6.1 Missing Values Analysis ............................................................. 54 4.2.6.2 Duplicate Records Detection ..................................................... 54 4.2.6.3 Data Type Verification .............................................................. 55 4.2.6.4 Outlier Detection ....................................................................... 55 vi 4.2.6.5 Constant Columns Check ........................................................... 56 4.2.6.6 Final Validation ......................................................................... 56 4.3 CTGAN Implementation ..................................................................................... 57 4.3.1 CTGAN Model Training .......................................................................... 57 4.3.2 Synthetic Data Generation ....................................................................... 57 4.3.3 Evaluation of Synthetic Data Quality ....................................................... 58 4.3.4 Evaluation Metrics for Synthetic Data ...................................................... 58 4.3.4.1 Mean (Average Value) .............................................................. 58 4.3.4.2 Standard Deviation .................................................................. 58 4.3.4.3 Minimum Value ........................................................................ 58 4.3.4.4 Maximum Value ....................................................................... 59 4.3.4.5 Distribution Plots ..................................................................... 59 4.3.4.6 Correlation Heatmap ................................................................. 59 4.3.4.7 Categorical Distribution Comparison ........................................ 59 4.3.4.8 Data Quality Score ....................................................................60 4.3.5 Importance of Visual Analysis ..................................................................60 4.4 Implementation Environment ..............................................................................60 4.4.1 Backend Technologies ..............................................................................60 4.4.2 Frontend Technologies ............................................................................. 61 4.4.3 Libraries and Tools ................................................................................... 61 4.4.4 Hardware and Development Setup ............................................................ 62 4.5 System Implementation ....................................................................................... 62 4.5.1 Project Structure ...................................................................................... 62 4.5.2 Implementation Snapshots ....................................................................... 63 4.5.2.1 Authentication Interface ............................................................ 63 4.5.2.2 Administrator Dashboard .......................................................... 64 4.5.2.3 Model Training and Management Interface ............................... 64 4.5.2.4 Synthetic Data Generation Interface ......................................... 65 4.5.2.5 Data Augmentation Interface .................................................... 65 4.5.2.6 Farmer Interfaces ...................................................................... 66 4.6 Evaluation and Validation Strategy ...................................................................... 67 4.6.1 Synthetic Data Evaluation (For Companies) ............................................ 67 4.6.2 Prediction Validation (For Farmers) ......................................................... 68 4.7 Conclusion .................................................................................... 69 General Conclusion 70 Bibliography |
| Type de document : | Mémoire master |
Disponibilité (1)
| Cote | Support | Localisation | Statut |
|---|---|---|---|
| MINF/969 | Mémoire master | bibliothèque sciences exactes | Consultable |




