Crop Yield Prediction and Efficient
use of
Fertilizers
T.Nikhil Yadav,Ram Chandra Prasad.
Abstract—Agriculture has been the backbone of the Indian economy and it will
continue to remain so for a long time. It has to support almost 17 per cent of world
population from 2.3 per cent of world geographical area.Data Mining is sky rocketing
in crop yield analysis.Every farmer is interested in knowing about the
yield he is about to get.This study will attempt to quantify past weather and
technological factors, analyze their effects on yields, predict future crop yields based
on projected levels of these factors and will also derive estimates of the parameters of
the random element associated with these yield predictions.The parameters are
loction,area ph value, percentage of nutrients like
potassium(K),nitrogen(N),phospurus(P) along with nutrient value of the soil in that
region,rainfall and use of API’s for weather,temperature,moisture is determined.Then
these parameters are analysed and the data will be trained with various suitable
machine learning algorithms for creating a stable model.This systems predicts crop
yield accurately and delivers end user with good valuable suggestions and the
required fertilizer ratio based on atmospheric and soil parameters.Which will help the
farmers to maximize their revenue and minimize the loss occured incorrect techniques
and wrong assumption of fertilizer usage.
Index terms:Neural network models,Back propogation
algorithm,Random forest tree algorithm,Accurate prediction.
I. Introduction
Agriculture is the primary source of livelihood for about 58 per cent of India’s
population. Gross Value Added by agriculture, forestry and fishing is estimated at Rs
18.55 lakh crore (US$ 265.51 billion) in FY19(PE).
The Indian food industry is poised for huge growth, increasing its contribution to world
food trade every year due to its immense potential for value addition, particularly
within the food processing industry. The Indian food and grocery market are the
world’s sixth largest, with retail contributing 70 per cent of the sales.
But now a days the weather conditions of the atmosphere are not stable like previous
days,because of global warming and other man made calamities.The weather
conditions are highly unstable and is very difficult for the farmers to rely upon the
traditional agricultural techniques.The farmers need more precise values for predicting
the wether conditions which in order gives the farmers a better crop yield.
Generally, machine learning algorithms will predict the most accurate output,of the
yield,farmrs use traditional methods and prior experience to predict.But we know the
weather conditions are higly unstable it is not possible for the farmer to predict
accurately.Therefore the technology will help the farmer to go for that vcrop or
not,and the neural models understand the pattern of the crop on various conditions
and predicts the type of the crop which farmer is going to crop.
There are many ways to make an algorithm for prediction and there are many
algorithms which are out there.But the main challenge is to get the most accurate
algorithm which has less error and does a precise job.
II. Related Work
The climate of India comprises a wide range of weather conditions across a vast
geographic scale and varied topography, making generalisations difficult. Based on
the Köppen system, India hosts six major climatic subtypes, ranging from arid deserts
in the west, alpine tundra and glaciers in the north, and humid tropical regions
supporting rainforests in the southwest and the island territories. Many regions have
starkly different microclimates, making it one of the most climatically diverse
countries in the world.Climate-related natural disasters cause massive losses of Indian
life and property. Droughts, flash floods, cyclones, avalanches, landslides brought on
by torrential rains, and snowstorms pose the greatest threats. Other dangers include
frequent summer dust storms, which usually track from north to south; they cause
extensive property damage in North India[55] and deposit large amounts of dust from
arid regions. Hail is also common in parts of India, causing severe damage to standing
crops such as rice and wheat.
Eswari et al 2018 [2] have indicated yield of the crop depends on the average,
minimum and maximum temperature and precipitation. Apart, from that, they have
taken one more attribute named crop evapotranspiration. The crop evapotranspiration
is a function of both the and growth stage and weather of the plant. This attribute is
taken into consideration to get a good decision on the yield of the groups.
S.V Baria [1] built up a strategy to demonstrate how a high- resolution satellite
imagery is essential to isolate rice cultivation. LAI’s multi-regression model was taken
as an input and NDVI or any other vegetation index calculated from red and near-
infrared spectral reflection was taken as another influence under normal
environmental conditions and common agronomic practices. The top methodology of
estimating rice yield using satellite imagery could be during the period of maximum
vegetative growth. SnehalS.Darikar et al. [2]discussed in their paper the use of
Artificial Neural Network to predict crop yield. The paper senses the parameters of the
regional soil and the various atmospheric conditions. Then it furthers analyses by
using feed forward back propagation ANN. By using Mat lab ANN approach was made
more efficient. They structure a system that accurately links climate effects to crop
yield, can be used to estimate long - term or short-term crop production and can also
obtain an ANN with adequate and useful data. AgaarwalSachin [3] discovered the air
quality index by utilizing neural system-based air quality factors, which worked with a
fractional number of informational indexes and are sufficiently strong to deal with
information with noise and inaccuracies. Various accessible varieties of neural system
models, for example, Repetitive Network Model (RNM), Consecutive System
Development Demonstrate (SNCM), Change Point Discovery Display with RNM (CPDM),
and Self Sorting out Element Maps (SOFM) are executed in this paper for anticipating
air quality. The created models were utilized to reproduce and figure the air quality
index dependent on long haul (yearly) and in addition present moment (every day)
informational collections. The models can accurately anticipate air quality patterns.
We develop a novel approach for augmenting parametric statistical models with deep
neural networks, which we term semiparametric neural networks (SNN). Used as a
crop yield modeling framework, the SNN achieves better out-of-sample predictive
performance than anything else yet published. By using prior knowledge about
important phenomena and the functional forms relating them to the outcome, the SNN
substantially improves statistical efficiency over typical neural networks. By
augmenting a parametric model with a neural network, it captures dynamics that are
either absent or imperfectly specified in parametric models. Specifically, we nest
linear-in-parameters regression specifications—taken from the yield and climate
modeling literature—within the top layer of the network (figure 1). Together with a
variety of complementary methods discussed below, this improves efficiency and
ultimately performance over both extant parametric approaches, and over fully-
nonparametric neural networks.
