By doing this project we aim to analyse different weather conditions for farming various crops and make clusters of the crops that grow in similar weather conditions. We aim to do multi class classification and prediction to figure out which crop can be grown with the given conditions and the clustering. This will help the farmers to produce better yield and varied range of various crops.
Hello, Agriculture accounts to about 68% of the population in India. The technological advancements have led agriculture’s share in GDP increase to 19.9% in 2020-2021 after 17 years. Although machines have been an instrumental factor driving technological aspects of agriculture, recently, data and its analysis has proved that it can be exploited to produce best results. Thousands of farmers are losing their lives to debts because they are not able to produce as much as before. Uneven rainfall, chemical imbalance in the soil contributes mainly to the decrease in yield. Thus our problem statement was developed.
We have gathered a dataset built by augmenting datasets of rainfall, climate and fertilizer data available from github. we hope this gives us a better idea of the trends of crops considering different environmental and geographical factors.
Applied descriptive statistics visualization and predictive modelling to identify the crops that can be grown at a stipulated temperature and other natural factors. While descriptive analysis enabled us to compare the Average Requirement for each crop at average conditions, hard clustering and various visualization techniques permitted us to identify homogeneous groups to develop an efficient mixed cropping strategy. Using regression we determined the strength and relationship b/w the dependent and series of independent variables. The accuracy of each model i.e; logistic regression, decision tree and SVM was hence calculated. Decision tree gave the highest accuracy so it was used for prediction.
IBM-Z was an interactive working console, with increased development capabilities and resource packages, that simplified our development process and improvised our analytical potentialities, producing the desired outputs. Using IBM-Z, the suitable crop to be produced for a given real time climatic condition has been predicted with an accuracy of 99% using accelerated and optimized analysis . IBM-Z allowed us to integrate multiple data together to analyse the crop patterns and cluster them using various algorithms. This allowed us to understand various trends in the crops and efficiently do mixed cropping and crop rotation. In the future, we plan to deploy it as an app making it possible to lend our hand to the farmers.
The Jupyter Notebook contains:
-
Descriptive Analysis of the Agriculture Data. i)checking summary of crops such as average of various parameters. ii)summary statistics of indivisual crop. using Interactive function to choose. iii)Average requirnment for each crop with given conditions.
-
Analysis of different agcriculture i)conditions using disc plot
-
categories the crop based on season clustering similar kind of crops using hard and soft clustering techniques.
-
Finding out any hiding Patterns in the Dataset.
-
Predictive Modeling.
