A Survey on Farm Productivity Prediction Using IoT
Pradeep Sudhakaran, C. Malathy, Ranjan Himanshu Ravi and Pranshu Sharma
The data collected from Google data sets of about 3977 is reduced and various algorithms are used such Decision tree classifier and random forest classifier is used. This dataset is divided into training and test data of 80 and 20 percent respectively. The soil type is found as one of the alkali, sandy, chalky, clay unpredictable changes in factors like rainfall and soil water content causes hard time for farmers. It can be improved with proper approach and analysis for the type of soil, warmth, max temp, min temp, humidity crop type And rainfall patterns. Crop and Weather prediction can be found by deriving useful inputs from these agricultural data available. Survey is conducted on the various algorithms used for climate, yield and productivity. The other challenge that our farmers face is land losing their productivity. Which crop should be planted based on the climate prediction and considering the strength of the soil keeping in mind the external factors like water availability and risk of crop to be destroyed due to pests. The efficiency of 99 percent is achieved by using random forest as machine learning tool. This is maximized performance.
Volume: Volume 24
Issues: Issue 1
Keywords: Farming, Internet of Things, Weather, Sensor and Productivity.