Crop Predictions Using Machine Learning

Guided By : Prof. M. P. Patil

Kalyani Prabhudesai1 ,Gayatri Salunkhe2,

Sanika Shinde3 ,Vaishnavi Suryavanshi4
1
Application Developer, Computer
Scienceand Engineering,D.Y.Patil
Technical Campus, Talsande, Maharashtra, India
E-mail: [email protected]
2
Application Developer, Computer
Scienceand Engineering,D.Y.Patil
Technical Campus, Talsande, Maharashtra,India.
E-mail: [email protected]
3
Testing Documentation, Computer
Science and Engineering, D.Y.Patil
Technical Campus, Talsande, Maharashtra,India
E-mail: [email protected]
4
Testing and Documentation,Computer
Science and Engineering, D.Y.Patil
TechnicalCampus,Talsande,Maharashtra,India
E-mail: [email protected]

Agriculture has an extensive history in India.

Abstract— India, with its predominantly agrarian Agriculture is one of the important occupation
economy, faces significant challenges in optimizing practiced in India. India being a nation of millions of
crop selection due to diverse climatic conditions, soil villages so it employs a large share of population in
variations, and fluctuating market demands. A large rural areas. It
segment of the farming community lacks access to
scientific methods and data-driven tools, resulting in
suboptimal crop yields and economic losses. This
is the broadest economic sector and plays a most
paper presents a machine learning-based crop
important role in the overall development of the
prediction system designed to assist farmers in
country. More than 60% of the land in the country is
making informed decisions about crop selection. By
used for agriculture in order to suffice the needs of 1.4
analyzing parameters such as soil composition,
billion people Thus adopting new agriculture
weather patterns, moisture levels, and historical yield
technologies is very important. Current agriculture is
data, the system recommends the most suitable crops
highly dependent on technology and focuses on
for a given field. The integration of machine learning
obtaining large profits from selected hybrid crops,
enables the identification of complex patterns and
which destroy the soil's physical and biochemical
correlations, enhancing the accuracy of predictions.
properties in the long run. This will be leads the
This approach not only improves crop productivity
farmers of our country towards profit.
and profitability but also supports sustainable
agricultural practices. The study underscores the Prior crop prediction and crop prediction was
importance of data science in modern farming and its performed on the basis of farmers experience on a
potential to transform traditional agricultural particular location. The crop is the significant factor
methods. The system aims to empower farmers with contributing in agricultural monetary. The crop
actionable insights, ultimately contributing to food depends on multiple factors such as climatic,
security and economic resilience in India. geographic, and financial elements. It is difficult for
farmers to decide when and which crops to plant.
I. INTRODUCTION
Farmers are unaware of which crop to grow, and what
 is the right time and place to start due to uncertainty MajorityVote(T1(x),T2(x),…,Tn(x))
in climatic conditions. They will prefer the prior or Where:
neighborhood or more trend crop in the surrounding • x={Soil pH,N,P,K,Temperature,Humidity,Rainfall}
region only for their land and they don't have enough • Ti(x)T_i(x)Ti(x): Prediction of the i-th decision tree
of knowledge about soil nutrients content such as in the forest
nitrogen, phosphorus, potassium in the land. • MajorityVote: The crop most frequently predicted by
Considering all these problems takes into the account all trees
we designed the system using a machine learning for
betterment of the farmer. Machine learning (ML) is a II. LITERATURE SURVE
game changer for agriculture sector.
Machine learning is the part of artificial Ashwani kumar Kushwaha[2] describes crop yield
intelligence, has emerged together with bigdata prediction methods and a suggest suitable crop so that
technologies and high- performance computing to it will improve the profit for the farmer and quality of
create new opportunities for data intensive science in the agriculture sector. In this paper for crop yield
the multidisciplinary agro technology domain. The prediction they obtain large volume data, it’s been
designed system will recommend the most suitable called as big data (soil and weather data) using
crop for particular land. Based on weather parameter Hadoop platform and agro algorithm. Hence based
and soil content such as Rainfall, Temperature, repository data will predict the suitability crop for
Humidity and pH. They are collected from particular condition and improvement crop quality.
Government website and KAGGLE. The system Girish L [3] describe the crop yield and rain fall
takes the required input prediction using a machine learning method. In this
from the farmers or data set such as Temperature, paper they gone through a different machine learning
Humidity and pH. This all inputs data applies to approaches for the prediction of rainfall and crop
machine learning predictive algorithms like logistic yield and also mention the efficiency of a different
regression and Decision tree to identify the pattern machine learning algorithm like liner regression,
among data and then process it as per input SVM, KNN method and decision tree. In that
conditions. The system recommends the crop for the algorithm they conclude that SVM have the highest
farmer and also recommends the amount of nutrients efficiency for rainfall prediction.
to be add for the predicted crop. Rahul katarya [4] describes the different machine
1. KEY-WORDS :- learning methods used for accelerating crop yield. In
this paper they gone through different artificial
• Machine Learning intelligence techniques such as machine learning
• Crop Recommendation algorithm, big data analysis for precision agriculture.
• Soil and Weather Data They explain about crop recommender system using
• Classification Algorithm KNN, Ensemble-based Models, Neural networks,
• Agricultural Analytics ...etc.

2. ALGORITHM :- III. PROPOSED SYSTEM

The Proposed system will predict the most suitable
• CNN Algorithm:-
crop for particular land based on soil contents and
A deep learning algorithm that uses convolutional
weather parameters such as Temperature, Humidity,
layers to automatically extract features from input soil PH and Rainfall.
data, especially effective for image and spatial data.

• Random Forest:-
An ensemble learning method that builds multiple
decision trees and merges their results to improve
accuracy and reduce overfitting.

Formula (Conceptual):
Crop_predicted=
 historical agricultural datasets. The model processes
the input parameters and predicts a crop index that
represents the most suitable crop for the current
conditions.
5. Crop Classification and Validation: The
predicted crop index is evaluated to verify its validity.
If the index corresponds to a known crop class, the
system proceeds to the next step. If the input
the prediction is invalid, the system displays an error
message indicating invalid readings.
1 Recommendation Display: Upon successful
classification, the system displays the most suitable
crop recommendation to the user. This enables the
farmer to make an informed, data-driven decision
regarding crop selection.
End: The system concludes the process after
Figure 1. Architecture of the proposed system. presenting the recommendation or the error message.

The Architecture of the proposed system consists

of various 3. Crop Prediction:
blocks as shown in the fig (1) as follows
A flowchart is simply a graphical representation of
The crop prediction system follows a systematic steps. It shows steps in sequential order and is widely
and structured workflow, as illustrated in Figure 1, to used in presenting the flow of algorithms, workflow
assist farmers in identifying the most suitable crop for or processes. Typically, a flowchart shows the steps as
cultivation based on environmental and soil boxes of various kinds, and their order by connecting
parameters. The major stages of the system are them with arrows. It originated from computer
outlined below: science as a tool for representing algorithms and
1. Start and Dataset Loading: The system initiates programming logic but had extended to use in all
by loading a comprehensive dataset that contains other kinds of processes. Nowadays, flowcharts play
historical agricultural data, including soil an extremely important role in displaying information
characteristics, temperature, humidity, and pH levels. and assisting reasoning. They help us visualize
complex processes, or make explicit the structure of
2. Data Pre-processing: In this stage, the dataset
problems and tasks. A flowchart can also be used to
undergoes cleaning and preparation. This includes
define a process or project to be implemented.
handling missing values, removing inconsistencies,
and normalizing data to ensure accuracy and quality.
Pre-processing is a crucial step to enhance the
performance of machine learning models.
3. Sensor Data Collection: The system reads real-
time environmental inputs such as temperature,
humidity, and soil pH values. These parameters are
essential for accurate crop recommendation, as they
directly influence crop growth and yield.
4. Prediction Using Machine Learning Algorithm:
The pre-processed data is fed into a machine
learning (ML) model, which has been trained on
 classification algorithms to assign a crop label to the
input set, effectively identifying which crop would
yield the best results under the given conditions.
To ensure the reliability of predictions, the system
includes a validation mechanism that checks whether
the input parameters fall within acceptable ranges. If
the values are valid, the system proceeds to classify
the crop and display the result. If the inputs are outside
the expected range, the system alerts the user by
displaying an error message indicating invalid
readings.
The accuracy of crop prediction is influenced by
the quality of data and the performance of the machine
learning model. Therefore, data preprocessing plays a
vital role in eliminating noise and inconsistencies.
Additionally, the system may incorporate ensemble
Figure 2. Flow chart learning techniques or multiple classifiers to improve
3.1. Crop Recommendation: prediction robustness.
Based on predicted rainfall, soil contents and Ultimately, the crop prediction feature provides
weather parameters the system will recommend the farmers with tailored, data-driven recommendations.
most suitable crop for cultivation. This system also This not only aids in optimal crop selection but also
provides details about required fertilizers like contributes to improved agricultural yield,
Nitrogen(N), Phosphorus (P) and potassium(K) in Kg profitability, and sustainability.
per hectare and display the required seed for a
cultivation in Kg per acre for recommended crop. IV. RESULT
This system as contain some other feature such as The crop prediction system is designed to
display the current market price and approximated recommend the most suitable crop for cultivation
yield in quintal per acre for recommended crop. Those based on key environmental and soil parameters such
all details will helps to farmers for choosing the most as temperature, humidity, soil pH, rainfall, and NPK
profitable crop. values. Annual rainfall is predicted using historical
data, while other parameters are provided by the user.
The system utilizes two machine learning
algorithms—Convolutional Neural Network
3.2.Crop Prediction: (CNN) and Random Forest—to analyze the data and
generate accurate crop recommendations. The CNN
The core functionality of the proposed system lies model captures complex data patterns, while the
in its ability to accurately predict the most suitable Random Forest algorithm offers high accuracy and
crop for cultivation based on environmental and soil- interpretability.
related parameters. This prediction is achieved
The output includes the recommended crop,
through the integration of machine learning
required seeds per acre, market price, approximate
algorithms that analyse various features including
yield, and the necessary NPK ratio. The comparative
temperature, humidity, and soil pH values. These
analysis of both algorithms helps in selecting the most
inputs are collected in real-time or derived from
efficient model for precise agricultural decision-
historical datasets and are essential for modelling the
making, ultimately assisting farmers in optimizing
complex relationships between crop requirements and
their crop selection and boosting productivity.
environmental conditions.
The prediction process begins with the system
receiving input values either through sensors or user
input. These values are then passed through a trained
machine learning model that has learned patterns
from historical agricultural data. The model utilizes
 have developed a farmer friendly system with GUI,
that will predict which would be the best suitable crop
for particular land and this system will also provide
information about required nutrients to add up,
required seeds for cultivation, expected yield and
market price. So, this makes the farmers to take right
Fig.:- Front page. decision in selecting the crop for cultivation such that
agricultural sector will be developed by innovative
idea.
FUTURE SCOPE:
We have to collect all required data by giving GPS
locations of a land and by taking access from Rain
forecasting system of by the government, we can
predict crops by just giving GPS location. Also, we
can develop the model to avoid over and under crisis
Fig. :- Register Page.
of the food.

REFERENCES
1. Prof. D.S. Zingade ,Omkar Buchade ,Nilesh Mehta
,Shubham Ghodekar ,Chandan Mehta “Crop
Prediction System using Machine Learning”.
2. Ashwani kumar Kushwaha, Swetabhattachrya “crop
yield prediction using agro algorithm in hatoop”.
3. Girish L, Gangadhar S, Bharath T R, Balaji K S,
Abhishek K T “Crop Yield and Rainfall Prediction in
Tumakuru District using Machine Learning”.
4. Rahul Katarya, Ashutosh Raturi, Abhinav
Mehndiratta, Abhinav
5. Thapper “Impact of Machine Learning Techniques in
Precision Agriculture”.
6. Pijush Samui, Venkata Ravibabu Mandla, Arun
Krishna and Tarun Teja “Prediction of Rainfall Using
Support Vector Machine and Relevance Vector
Machine”.
7. Himani Sharma, Sunil Kumar “A Survey on Decision
Tree
8. Algorithms of Classification in Data Mining”.
9. Pavan Patil, Virendra Panpatil, Prof. Shrikant Kokate
“Crop Prediction System using Machine Learning
Algorithms”.

Fig. Result
IIV. CONCLUSION AND FUTURE SCOPE

CONCLUSION:
Presently our farmers are not effectively using
technology and analysis, so there may be a chance of
wrong selection of crop for cultivation that will
reduce their income. To reduce those type of loses we