Predicting Agricultural Output by Applying Machine Learning

Author(s):

Siddhant Ghosh , Suman Ramesh Tulsiani Technical Campus, SPPU; Prof. Sonali Patil, Suman Ramesh Tulsiani Technical Campus, SPPU; Utkarsha Matere, Suman Ramesh Tulsiani Technical Campus, SPPU

Keywords:

Gaussian Regression, KNN, Naïve Bayes Clustering, Machine learning, K-Means

Abstract:

Data sets of different related to the agriculture is vastly available. The main motto of our system will be to create a machine learning algorithm and the use the available datasets for the unsupervised machine learning process and thereby predict the future crop production, crops which would likely to give a greater profit margin, etc. The prediction can also help various State as well as Central Government for taking appropriate steps as described by the prediction. The methodology will be as follows: By taking various datasets from the government and by the help of clustering for the first step and in the next step linear regression for correctly predicting. The reason why we are using unsupervised learning for the first step is so that we won’t lose any data. The whole system will be available as a website with a GUI that can accept data for the learning process. The output will also depend upon the present state of the environment. The Features for the Learning Process are:  Weather Data (Rainfall, Winds, Frequency of Droughts, so on.)  Soil Composition.  Types of Fertilizers used.  The Source for Seeds/Saplings  Seasonal Data.  Whether mechanized farms or manual labour used. (If manual labour is not used what is extent of mechanization).

Other Details:

Manuscript Id	:	IJSTEV4I11053
Published in	:	Volume : 4, Issue : 11
Publication Date	:	01/06/2018
Page(s)	:	86-100

Download Article