Research and Analysis of the Application of Machine Learning in Agricultural Development

Agriculture is the most basic, fundamental and important industry. Now, amid global climate change and resource shortages, agriculture must deal with the challenges of growing demand as the world's population increases This article organizes three aspects of agriculture that need improvement: anticipatory preparation before production, improvement of production methods, and detection and classification of agricultural products, and analyzes how machine learning can help agricultural progress in these three aspects. Residual deep convolution and spatial pyramid pooling algorithms in machine learning can be used to help detect plant pests and diseases. The RF algorithm, XGBoost algorithm, LightGBM algorithm and CatBoos in machine learning can generate landslide susceptibility maps. Deep learning, convolutional neural networks, and support vector machines can identify hybrid wheat. Through this research, it can be determined that machine learning can be of great help to agricultural development, and this help and development is mutual. The significance of this study lies in how machine learning can help agricultural development and face these problems.