Prediction of crop harvest soil composition using vector distance analysis And multi linear regression

For many decades, agriculture is involved in high-quality of crop productions and plays vital role in the world's economy. Many factors affect the agriculture in direct and indirect ways. The optimization of agriculture production is implemented in many ways like finding the appropriate location for the cultivation of crops and forecasting the harvest based on the factors in farm fields etc., It is calculated by collecting the past crop data from the fields and data sets are generated using the discrete data, which provides significant information to increase the harvest of crops. The proposed system is aimed at improving the prediction accuracy. We have implemented “Single Prediction Algorithm” in order to increase the accuracy by pruning the unwanted data and extracting only the valuable data. The accurate preferred factor is calculated by applying vector distance method. Then the multi linear regression is applied on the newly generated data and old collected data simultaneously to better predict the crop harvest. By simply improving the data set and producing new improved values from the existing data set the multi linear regression is able to predict with high accuracy. The humidity in the soil is taken as one of the factor in the agriculture. The weather elements are also taken in to account with the help of temperature parameter. The semantic processing of the historical data predicts the humidity, weather, temperature and growth rate of the crops and plants. The prediction of these parameters detects the exact growth rate in the available rates. The sematic processing of features selected helps to identify exact numerical values which are further used for liner regression techniques. The percentage of growth rate is 95% in the case of prediction and recommendation of manipulating the historical values.