This paper presents a network intrusion detection system (NIDS) designed to enhance the security of smart agriculture environments, which are increasingly vulnerable to cyber threats due to the integration of IoT technologies. The authors developed a framework using radial basis function neural networks (RBFNN) to detect and classify intrusions in IoT networks. To improve the model's performance, they applied crowd wisdom tree-based machine learning techniques, including Random Forest (RF), AdaBoost (ADA), Extra Trees (ET), LightGBM (LGBM), and XGBoost (XGB), to select relevant features from the datasets. A single-class support vector machine (1- CSVM) was used to detect and remove outliers. The model was evaluated using NF-Bot-IoT and NF-ToN-IoT datasets, achieving 99.25% accuracy and 82.97% Matthews correlation coefficient (MCC) on the NF-Bot-IoT dataset, and 90.05% MCC and 96.92% ACC on the preprocessed NF-ToN-IoT dataset. The model demonstrated excellent performance, particularly in handling the imbalance of the NF-Bot-IoT dataset. The paper's main contributions are the integration of multiple tree-based machine learning techniques for feature selection and the application of 1- CSVM for outlier detection. The authors also evaluated the model's performance on two datasets, showing its effectiveness in addressing dataset imbalances. The remainder of the paper includes a literature review, a detailed description of the proposed approach, experimental results, and a conclusion.This paper presents a network intrusion detection system (NIDS) designed to enhance the security of smart agriculture environments, which are increasingly vulnerable to cyber threats due to the integration of IoT technologies. The authors developed a framework using radial basis function neural networks (RBFNN) to detect and classify intrusions in IoT networks. To improve the model's performance, they applied crowd wisdom tree-based machine learning techniques, including Random Forest (RF), AdaBoost (ADA), Extra Trees (ET), LightGBM (LGBM), and XGBoost (XGB), to select relevant features from the datasets. A single-class support vector machine (1- CSVM) was used to detect and remove outliers. The model was evaluated using NF-Bot-IoT and NF-ToN-IoT datasets, achieving 99.25% accuracy and 82.97% Matthews correlation coefficient (MCC) on the NF-Bot-IoT dataset, and 90.05% MCC and 96.92% ACC on the preprocessed NF-ToN-IoT dataset. The model demonstrated excellent performance, particularly in handling the imbalance of the NF-Bot-IoT dataset. The paper's main contributions are the integration of multiple tree-based machine learning techniques for feature selection and the application of 1- CSVM for outlier detection. The authors also evaluated the model's performance on two datasets, showing its effectiveness in addressing dataset imbalances. The remainder of the paper includes a literature review, a detailed description of the proposed approach, experimental results, and a conclusion.