Machine Learning (ML) plays a crucial role as a decision-support tool for Crop Yield Prediction (CYP), aiding in key agricultural decisions such as crop selection and management strategies throughout the growing season. The primary goal of crop yield estimation is to enhance agricultural productivity, achieved through various well-established predictive models.

The Power Transformations block, particularly the Yeo-Johnson method, is essential in the preprocessing phase of a machine learning pipeline for crop yield prediction. Many ML algorithms assume input features follow a normal distribution, but real-world agricultural data is often skewed due to factors like climate variability, soil diversity, and farming practices. The Yeo-Johnson transformation helps normalize such data, improving model accuracy and reliability. 

This paper presents the Smart Agriculture Yield and Fertilizer Optimization System (SAYFOS), an innovative solution addressing key agricultural challenges. SAYFOS leverages advanced data analytics, IoT technologies, and machine learning to enable real-time crop health and soil condition monitoring. By integrating data from soil sensors, weather forecasts, and satellite imagery, SAYFOS optimizes fertilizer application rates and enhances crop yield predictions with high accuracy, ensuring smarter and more sustainable farming. 

The Smart Irrigation Yield Optimization (SIYO) System is a cutting-edge IoT-based solution designed to enhance crop yield and growth predictions. Traditional irrigation methods—such as furrow irrigation, overhead sprinklers, and manual watering—often suffer from inefficiencies in precision and resource management. SIYO addresses these challenges by deploying a network of real-time sensors that monitor soil moisture, temperature, humidity, and nutrient levels. This enables automated, data-driven irrigation, ensuring crops receive the right amount of water and nutrients at the right time. By adapting to changing environmental conditions, SIYO optimizes resource use while maximizing crop health and productivity. 

This research focuses on developing a machine learning-based model for accurate crop yield prediction. The proposed model utilizes advanced techniques such as Inception ResNet V2, AlexNet, Random Forest, Decision Tree, KNN, Naive Bayes, Simple RNN, Logistic Regression, and a Voting Classifier to analyze historical data and identify key yield-influencing factors. By incorporating variables like land area, crop variety, and climatic conditions, the model offers data-driven insights, empowering farmers to make informed decisions for improved productivity.

This research paper proposes the Crop Yield Prediction Model (CRY), which employs an adaptive cluster approach on a dynamically updated historical crop dataset to enhance precision agriculture decision-making. CRY utilizes a bee hive modeling approach to analyze and classify crops based on growth patterns and yield trends. The classified dataset has been tested using Clementine, integrating existing crop domain knowledge to improve prediction accuracy and agricultural planning.