CROP PREDICTING SYSTEM USING MACHINE LEARNING

A project Report submitted

in partial fulfillment for the award of the Degree of

Bachelor of Technology
in
Computer Science and Engineering
by

VIKASH KUMAR SINGH (U18CS021)

HIMANSHU KAUSHIK (U18CS042)
CHINDAM ROHITHKUMAR (U18CS054)
KOPPU BHARATH (U18CS055)

Under the guidance of

Mrs. C. ANURADHA

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

SCHOOL OF COMPUTING

BHARATH INSTITUTE OF HIGHER EDUCATION AND RESEARCH

(Deemed to be University Estd u/s 3 of UGC Act, 1956)

CHENNAI 600073, TAMILNADU, INDIA

April , 2022

i
 CERTIFICATE

This is to certify that the project report entitled CROP PREDICTING

SYSTEM USING MACHINE LEARNING submitted by Vikash Kumar Singh
(U18CS021), Himanshu Kaushik (U18CS042), Chindam Rohithkumar (U18CS054),
Koppu Bharath (U18CS055) to the Department of Computer Science and Engineering,
Bharath Institute of Higher Education and Research, in partial fulfillment for the
award of the degree of B. Tech in Computer Science and Engineering is a bonafide
record of project work carried out by them under my supervision. The contents of this
report, in full or in parts, have not been submitted to any other Institution or University
for the award of any other degree.

------------------------
Mrs. C. Anuradha
Supervisior
Department of Computer Science & Engineering,
School of Computing,
Bharath Institute of Higher Education and Research
April, 2022

-------------------------------
Dr. B. Persis Urbana Ivy
Professor & Head
Department of Computer Science & Engineering,
School of Computing,
Bharath Institute of Higher Education and Research,
April, 2022

INTERNAL EXAMINER EXTERNAL EXAMINER

ii
w

DECLARATION

We declare that this project report titled CROP PREDICTING

SYSTEM USING MACHINE LEARNING submitted in partial fulfillment
of the degree of B. Tech in Computer Science and Engineering is a
record of original work carried out by us under the supervision of
Mrs. C. Anuradha , and has not formed the basis for the award of any other
degree or diploma, in this or any other Institution or University. In keeping
with the ethical practice in reporting scientific information, due
acknowledgements have been made wherever the findings of others have been
cited.

--------------------------
Vikash Kumar Singh
U18CS021

-----------------------
Himanshu Kaushik
U18CS042

----------------------------
Chindam Rohithkumar
U18CS054

-------------------
Koppu Bharath
U18CS055

Chennai
-- / -- / ----

iii
 ACKNOWLEDGMENTS

First, we wish to thank the almighty who gave us good health and success throughout our
project work.

We express our deepest gratitude to our beloved President Dr. J. Sundeep Aanand, and
Managing Director Dr. E. Swetha Sundeep Aanand for providing us the necessary facilities for
the completion of our project.
We take great pleasure in expressing sincere thanks to Vice Chancellor (I/C) Dr. K. Vijaya
Baskar Raju, Pro Vice Chancellor (Academic) Dr. M. Sundararajan, Registrar Dr. S.
Bhuminathan and Additional Registrar Dr. R. Hari Prakash for backing us in this project.
We thank our Dean Engineering Dr. J. Hameed Hussain for providing sufficient facilities
for the completion of this project.
We express our immense gratitude to our Academic Coordinator Mr. G. Krishna Chaitanya
for his eternal support in completing this project.

We thank our Dean, School of Computing Dr. S. Neduncheliyan for his encouragement
and the valuable guidance.

We record indebtedness to our Head, Department of Computer Science and

Engineering Dr. B. Persis Urbana Ivy for immense care and encouragement towards us
throughout the course of this project.

We express our immense gratitude to our Project Coordinator Dr. Anitha Karthi for
his eternal support in completing this project.

We also take this opportunity to express a deep sense of gratitude to our Supervisor Mrs.
C. Anuradha for her cordial support, valuable information and guidance, she helped us in
completing this project through various stages.

We thank our department faculty, supporting staff and friends for their help and
guidance to complete this project.
VIKASH KUMAR SINGH (U18CS021)

HIMANSHU KAUSHIK (U18CS042)

CHINDAM ROHITHKUMAR (U18CS054)

KOPPU BHARATH (U18CS055)

iv
 ABSTRACT

As we know the fact that, India is the second largest population country in the
world and majority of people in India have agriculture as their occupation. Some of
farmers are facing crop selecting problem for their field. If farmer wants to give good
protein vegetables than farmers need to select the crop according to their soil
requirements. So, we have designed the system using machine learning algorithms for
betterment of farmers. Our system will suggest the best suitable crop for particular land
based on content and weather parameters. Hence by utilizing our system farmers may
increase in profit margin and can give good protein vegetables. Data science have
proved that data which we have plays a vital role in predictions. Data science in
agriculture is a growing field and has a wide scope in future. The prediction made by
machine learning algorithms will help the farmers to come to a decision which crop to
grow to induce the most suitable crop by considering factors like temperature, rainfall,
soil parameters, etc.

v
 TABLE OF CONTENT
DESCRIPTION PAGE NUMBER

CERTIFICATE ii
DECLARATION iii
ACKNOWLEDGEMENTS iv
ABSTRACT v
LIST OF FIGURES viii
LIST OF TABLES ix
ABBREVIATIONS/ NOTATIONS/ NOMENCLATURE x
1. INTRODUCTION 1
1.1. CROP PREDICTION 1
1.2. OBJECTIVE 2
1.3. ORGANIZATION OF REPORT 2
2. LITERATURE SURVEY 3
3. EXISTING SYSTEM & PROPOSED SYSTEM 6
3.1. EXISTING SYSTEM 6
3.2. PROPOSED SYSTEM 6
4. IMPLEMENTATION 8
4.1. DATA COLLECTION 8
4.2. DATA PREPROCESSING 8
4.3. MACHINE LEARNING ALGORITHMS 9
4.4. CROP PREDICTION 10
4.5. DEPLOYMENT 10
5. METHODOLOGY 13
5.1. BASIC PROCESS 13
5.2. DATA SETS 13
5.3. EXPLORATORY DATA ANALYSIS 14
5.4. EDA PERFORMED 15
5.5. ALGORITHM USED 16
5.6. RANDOM FOREST 17
6. SYSTEM ANALYSIS AND DESIGN 19
6.1. SYSTEM ARCHITECTURE 19
6.2. FLOWCHART 20
7. SYSTEM REQUIREMENTS SPECIFICATIONS 22
7.1. BASIC REQUIREMENTS 23
7.2. REQUIREMENT 24

vi
 7.2.1. HARDWARE REQUIREMENTS 24
7.2.2. SOFTWARE REQUIREMENTS 26
8. RESULT AND DISCUSSION 30
8.1. RESULT 30
9. CONCLUSION & FUTURE WORK 32
9.1. CONCLUSION 32
9.2. FUTURE WORK 32
REFERENCES 33
APPENDIX 34

vii
 LIST OF FIGURES
TITLE PAGE NUMBER

4.1. ARCHITECTURE OF CROP PREDICTION 8

4.2. STRUCTURE OF DECISION TREE 9

4.3. WORKING OF RANDOM FOREST 10

4.4. IMPEMENTING FLASK 11

4.5. TRAIN & TESTING MODEL 11

4.6. PREDICTION CODE 12

4.7. CREATED WEB PAGE 12

5.1. GENRAL PROCESS 13

5.2. SOIL & CROP DATA SAMPLE 14

5.3. CORERELATION MATRIX EXAMPLE 15

5.4. EDA CODE 15

5.5. OUTPUT OF DESCRIBE FUNCTION 15

5.6. CORERELATION MATRIX OF THE PROPOSED SYSTEM 16

5.7. RANDOM FOREST FLOW 17

6.1. SYSTEM ARCHITECTURE 20

6.2. RESEARCH OVERVIEW 21

7.1. PROCESSOR 24

7.2. WIFI 25

7.3. HARD DRIVE 25

7.4. RAM 26

7.5. JUPYTER NOTEBOOK 27

7.6. PYTHON 27

7.7. PYCHARM 28

7.8. FLASK 28

7.9. VISUAL STUDIO 29

viii
 LIST OF TABLES

TITLE PAGE NUMBER

2.1. APPROACH TO CROP PREDICTION 5

5.1. SUMMARY OF THE APPROACHES 17

8.1. TABLE OF ACCURACY LEVELS 30

ix
 ABBREVIATION/ NOTATIONS/ NOMENCLATURE

ML MACHINE LEARNING

RF RANDOM FOREST

DT DECISION TREE

K POTASSIUM

P PHOSPHORUS

N NITROGEN

CSS CASCADING STYLE SHEETS

HTML HYPER TEXT MARKUP LANGUAGE

EDA EXPLORATORY DATA ANALYSIS

PH POTENTIAL OF HYDROGEN

SRS SYSTEM REQUIREMENT SPECIFICATIONS

x
 CHAPTER 1
INTRODUCTION

1.1. CROP PREDICTION

Agriculture has an extensive history in India. Agriculture is one of the important occupation
practiced in India. India being a nation of millions of villages so it employs a large share of population
in rural areas. It is the broadest economic sector and plays a most important role in the overall
development of the country. More than 60% of the land in the country is used for agriculture in order
to suffice the needs of 1.4 billion people Thus adopting new agriculture technologies is very important.
Current agriculture is highly dependent on technology and focuses on obtaining large profits from
selected hybrid crops, which destroy the soil's physical and biochemical properties in the long run. This
will be leads the farmers of our country towards profit.

Prior crop prediction and crop prediction was performed on the basis of farmers experience on a
particular location. The crop is the significant factor contributing in agricultural monetary. The crop
depends on multiple factors such as climatic, geographic, and financial elements. It is difficult for
farmers to decide when and which crops to plant. Farmers are unaware of which crop to grow, and what
is the right time and place to start due to uncertainty in climatic conditions .They will prefer the prior
or neighborhood or more trend crop in the surrounding region only for their land and they don’t have
enough of knowledge about soil nutrients content such as nitrogen, phosphorus, potassium in the land.

Considering all these problems takes into the account we designed the system using a machine
learning for betterment of the farmer. Machine learning (ML) is a game changer for agriculture sector.
Machine learning is the part of artificial intelligence, has emerged together with bigdata technologies
and high-performance computing to create new opportunities for data intensive science in the multi-
disciplinary agro technology domain.

The designed system will recommend the most suitable crop for particular land. Based on weather
parameter and soil content such as Rainfall, Temperature, Humidity and pH. They are collected from
Government website and KAGGLE. The system takes the required input from the farmers or data set
such as Temperature, Humidity and pH. This all inputs data applies to machine learning predictive
algorithms like logistic regression and Decision tree to identify the pattern among data and then process
it as per input conditions. The system recommends the crop for the farmer and also recommends the
amount of nutrients to be add for the predicted crop.

1
1.2. OBJECTIVE

• Contributing to optimal crop growth, development and yield.

• Predict appropriate crop from given temperature and rainfall and soil.
• To enhance the economic development of all stake-holders.
• To improve nutritional standards for betterment of health.
• To contribute towards protection and up gradation of the environment for ensuing
ecological balance, avoidance of global warming and healthy living for man and animal.
• to reduce the financial losses faced by the farmers caused by planting the wrong crops
• Also to help the farmers to find new types of crops that can be cultivated in their area. So,
to make every farmer rich by farm and wealth we producing this System.

1.3. ORGANIZATION OF REPORT

The following chapter describes literature survey. Chapter 3 gives a overview about system
requirements of this project. Chapter 4 provides technical details about System analysis and design.
Chapter 5 and 6 describes System Implementation and System Methodology. Chapter 7 and Chapter
8 gives the Existing System, Proposed System and Result of this project. Chapter 9 will give the
conclusion and future work of our system and future work.

2
 CHAPTER 2
LITERATURE SURVEY
Given the significance of crop prediction, numerous suggestions have been proposed in the past
with the goal of improving crop prediction accuracy. In this paper different machine learning
methodology has been approached to model and predict various crop yields at rural areas based on
parameters of soil(PH, nitrogen, potassium, etc.) and parameters related to the atmosphere (rainfall,
humidity, etc.)

It looks at five of Tamil Nadu's most important crops- rice, maize, ragi, sugarcane, and tapioca
during a five-year period beginning in 2005. In order to get the maximum crop
productivity, various factors such as rainfall, groundwater, and cultivation area, and soil type were
used in the analysis. K-Means technique was used for the clustering, and for the classification, the
study looked at three different types of algorithms: fuzzy, KNN, and Modified KNN. After the
analysis, MKNN gave the best prediction result of the three algorithms.

An application for farmers can be created that will aid in the reduction of many problems in the
agriculture sector. In this application, farmers perform single/multiple testing by providing input such
as crop name, season, and location. As soon as one provides the input, the user can choose a method
and mine the outputs. The outputs will show you the crop's yield rate. The findings of the
previous year's data are included in the datasets and transformed into a supported format. The machine
learning models used are Naïve Bayes and KNN.

To create the dataset, information about crops over the previous ten years was gathered from
a variety of sources, such as government websites. An IoT device was setup to collect the atmospheric
data using the components like Soil sensors, Dht11 sensor for humidity and temperature, and Arduino
Uno with Atmega as a processor. Naive Bayes, a supervised learning algorithm obtaining an accuracy
of 97% was further improved by using boosting algorithm, which makes use of weak rule by an
iterative process to bring higher accuracy.

To anticipate the yield, the study employs advanced regression techniques such as ENet,
Kernel Ridge, and Lasso algorithms. The three regression techniques are improved by using Stacking
Regression for better prediction. However, when a comparison study is conducted between the existing
system and the proposed system employing Naive Bayes and Random Forest, respectively. The
proposed system comes out on top. Because it is a bagging method, the random forest algorithm has a

3
high accuracy level, but the Naïve Bayes classifier’s accuracy level is lower as the algorithm is
probability based.

It contributes to the following aspects- (a) Crop production prediction utilizing a range of
Machine Learning approaches and a comparison of error rate and accuracy for certain regions. (b) An
easy-to-use mobile app that recommends the most gainful crop. (c) A GPS-based location identifier that
can be used to obtain rainfall estimates for a specific location. (d) A system that recommends the prime
time to apply fertilizers. On the given datasets from Karnataka and Maharashtra, different machine
learning algorithms such as KNN, SVM, MLR, Random Forest, and ANN were deployed and assessed
for yield to accuracy. The accuracy of the above algorithms is compared. The results show that Decision
Tree is the most accurate of the standard algorithms used on the given datasets, with a 99.87% accuracy
rate.

Regression Analysis is applied to determine the relationship between the three factors: Area
Under Cultivation, Food Price Index, and Annual Rainfall and their impact on crop yield. The above
three factors are taken as independent variables, and for the dependent variable, crop yield is taken
into consideration. The R2 obtained after the implementation of RA shows these three factors showed
slight differences indicating their impact on the crop yield.

In the proposed paper, the dataset is collected from the government websites such as
APMC website, VC Farm Mandya, which contains data related to climatic conditions and soil nutrients.
Two machine learning models were used; the model was trained using the Support Vector Machine
model with Radial Basis Function kernel for rainfall prediction and Decision Tree for the crop
prediction.

A comparative study of various machine learning can be applied on a dataset with a view
to determine the best performing methodology. The prediction is found by applying the Regression
Based Techniques such as Linear, Random Forest, Decision Tree, Gradient Boosting, Polynomial and
Ridge on the dataset containing details about the types of crops, different states, and climatic
conditions under different seasons.

The parameters used to estimate the efficiency of these techniques were mean absolute
error, root mean square error, mean squared error, R-square, and cross validation. Gradient Boosting
gave the best accuracy- 87.9% for the target variable ‘Yield’ and Random Forest- 98.9% gave
the best accuracy for the target value ‘Production’. The DHT22 sensor is recommended for
monitoring live temperature and humidity.

4
 The surrounding air is measured with a thermistor and a capacitive humidity sensor and outputs
a digital signal on the data pin to the Arduino Uno port pin. The humidity value ranges from 0-100%
RH and -40 to 80 degrees Celsius to read the temperature. The above two parameters and soil
characteristics are considered as input to three different machine learning models: Support Vector
Machine, Decision Tree, and KNN. The Decision Tree gave better accuracy results.

Table 2.1: Approach to Crop Prediction

5
 CHAPTER 3
EXISTING SYSTEM & PROPOSED SYSTEM

3.1 Existing System

The crop yield rate plays a significant role in the economic growth of the country. So, there is
a need to increase crop yield rate. Some biological approaches (e.g. seed quality of the crop, crop
hybridization, strong pesticides) and some chemical approaches (e.g. use of fertilizer, urea, potash) are
carried out to solve this issue. One of existing system we identified is Crop Selection Method (CSM) to
achieve a net yield rate of crops over the season. We have taken example of CSM to demonstrate how
it helps farmers in achieving more yield Crop.
These can be classified as:

• Seasonal crops—crops can be planted during a season. e.g. wheat, cotton.

• Whole year crops—crops can be planted during the entire year. e.g. vegetable, paddy, Toor.
• Short time plantation crops— crops that take a short time for growing. e.g. potato, vegetables,
ratio.
• Long-time plantation crops— These crops take a long time for growing. e.g. sugarcane, Onion.
A combination of these crops can be selected in a sequence based on yield rate per day. Illustrates
sequences of crops with cumulative yield rate over the season. CSM method, shown in may improve
the net yield rate of crops using the limited land resource and also increases re-usability of the land.
The agricultural systems that significantly follows to the agriculture of India are subsistence
farming, organic farming, industrial farming. Regions all over India differ in types of farming they use;
some are based on horticulture, ley farming, agroforestry, and many more. The surveyed research papers
have given a rough idea about using ML with only one attribute. We have the aim of adding more
attributes to our system and ameliorate the results, which can improve the yields and we can recognize
several patterns for predictions. This system will be useful to justify which crop can be grown in a
particular region.

3.2 Proposed System

The Proposed system will predict the most suitable crop for particular land based on soil
contents and weather parameters such as Temperature, Humidity, soil PH and Rainfall. The
Architecture of the proposed system consists of various blocks such as:

Data Collection: - Data collection is the most efficient method for collecting and measure the data
from different resources like govt websites, VC Form Mandya, APMC website …. etc. To get an
approximate dataset for the system. This dataset must contain the following attributes i) Soil PH ii)
Temperature iii) Humidity iv) Rainfall v) Crop data vi) NPK values, those parameters will consider for
crop prediction.

Data Preprocessing: - After collecting datasets from various resources. Dataset must be
preprocessing before training to the model. The data preprocessing can be done by various stages,

6
begins with reading the collected dataset the process continues to data cleaning. In data cleaning the
datasets contain some redundant attributes, those attributes are not considering for crop prediction. So,
we have to drop unwanted attributes and datasets containing some missing values we need to drop these
missing values or fill with unwanted nan values in order to get better accuracy.

Machine Learning Algorithm for Prediction: - Machine learning predictive algorithms has
highly optimized estimation has to be likely outcome based on trained data.

Crop Recommendation: Based on predicted rainfall, soil contents and weather parameters the
system will recommend the most suitable crop for cultivation. Crop prediction process being with the
loading the external crop datasets. Once the dataset read then pre-processing will be done by various
stages as discussed in Data Pre-processing section. After the data pre-processing, train the models using
Decision tree classifier into training set. For a prediction of the crop, we consider a various factor such
as temperature, humidity, soil PH and predicted rainfall. Those are the input parameter for a system that
can be entered by manually or taken from the sensors. Predicted rainfall and input parameter values will
be appended in a list. The Decision tree algorithm will predict the crop based on list data.
START

Read temperature, Humidity, pH Values

Append all the values in list & Load External Crop dataset

DATA PREPROCESSING
(dealing with missing values, data cleaning, train/test split)

YES

Predict Crop Using Decision Tree Regressor Crop Classification and by using NO Predicted crop

index values

YES

7
 Display Suitable Crop

Chapter 4
Implementation

The system will predict the most suitable crop for particular land based on soil parameters and
weather parameters such as Temperature, Humidity, soil PH, Potassium(K), Phosphorus(P),
Nitrogen(N) and Rainfall.

Fig.4.1: Architecture of crop prediction

4.1. Data Collection: -

Data collection or Loading Data is the most efficient method for collecting and measure the data
from different resources like govt. websites, Kaggle website .... etc. To get an approximate dataset for
the system. This dataset must contain the following attributes i)Soil Ph ii) Temperature iii) Humidity
iv) Rainfall v) Crop data vi) NPK values, those parameters will consider for crop prediction.

4.2. Data Preprocessing: -

After collecting datasets from various resources. Dataset must be preprocessing before training to
the model. The data preprocessing can be done by various stages, begins with reading the collected

8
dataset the process continues to data cleaning. In data cleaning the datasets contain some redundant
attributes, those attributes are not considering for crop prediction. So, we have to drop unwanted
attributes and datasets containing some missing values we need to drop these missing values or fill with
unwanted nan values in order to get better accuracy. Then define the target for a model. After data
cleaning the dataset will be split into training.

4.3. Machine Learning Algorithm for Prediction: -

Machine learning predictive algorithms has highly optimized estimation has to be likely outcome
based on trained data. Predictive analytics is the use of data, statistical algorithms and machine learning
techniques to identify the likelihood of future outcomes based on historical data. The goal is to go
beyond knowing what has happened to providing a best assessment of what will happen in the future.
First, In our system we check the maximum accuracy between some algorithms such as Logistic
Regression, Decision Tree and Random Forest.

4.3.1 Logistic Regression:

Logistic regression is one of the most popular Machine Learning algorithms, which comes
under the Supervised Learning technique. It is used for predicting the categorical dependent
variable using a given set of independent variables. Logistic regression predicts the output of a
categorical dependent variable.

4.3.2 Decision Tree:

It is a supervised learning algorithm where attributes and class labels are represented using a
tree. Here, root attributes are compared with the record's attribute and subsequently, depending
upon the comparison, a new node is reached. This comparison is continued until a leaf node with
a predicted class value is reached. Therefore, a modeled decision tree is very efficient for prediction
purposes

9
 Fig. 4.2: Structure of decision tree

4.3.3 Random Forest:

It is an ensemble method of learning that is commonly used for both classification and
regression. In order to train the model to perform prediction using this algorithm, the test features
must be passed through the rules of each randomly created tree. As a result of this, a different target
will be predicted by each random forest for the same test feature. Then, votes are calculated on the
basis of each predicted target. The final prediction of the algorithm is the highest votes predicted
target. The fact that random forest algorithm can efficiently handle missing values and that the
classifier can never over-fit the model are huge benefits for using this algorithm.

Fig.4.3: Working Of Random Forest

We will discuss about accuracy levels of the above three algorithms while Result section.

4.4. Crop Prediction:

Crop prediction process being with the loading the external crop datasets. Once the dataset read then
pre-processing will be done by various stages as discussed in Data Pre-processing section. For a
prediction of the crop, we consider a various factor such as temperature, humidity, soil PH, Soil
Parameters and rainfall. Those are the input parameter for a system that can be entered by manually.
The most accuracy algorithm among the three Logistic Regression, Decision Tree and Random Forest
will predict the crop based on data.

4.5. DEPLOYMENT
The proposed system recommends the best suitable crop for particular land by considering
parameters as annual rainfall, temperature, humidity and soil PH.

10
 In order to deploy the trained model for the farmers to use off, we would need a web application
with a simple user interface which the farmers can utilize.

Fig.4.4: Implementing Flask

Fig.4.5: Train & Testing Model

11
 Fig.4.6: Prediction Code

Thus, here we made a simple web interface using HTML & CSS.

Fig.4.7: Created Web Page

12
 CHAPTER 5
METHODOLOGY

5.1 Basic Process

i. Data Collection: Collect the data that the algorithm will learn from.
ii. Data Preparation: Format and engineer the data into the optimal format, extracting
important features and performing dimensionality reduction.
iii. Training: Also known as the fitting stage, this is where the Machine Learning algorithm
actually learns by showing it the data that has been collected and prepared.
iv. Evaluation: Test the model to see how well it performs.
v. Tuning: Fine tune the model to maximize its performance.

\
Fig.5.1: General Process

5.2 Datasets
Machine Learning depends heavily on data. It‘s the most crucial aspect that makes algorithm
training possible. It uses historical data and information to gain experiences. The better the collection
of the dataset, the better will be the accuracy.
The first step is Data Collection. For this project, we require two datasets. One for modelling
the crop prediction algorithm and other for predicting weather .i.e. Average Rainfall and Average
Temperature. These two parameters are predicted so as to be used as inputs for predicting the suitable
crop.
The crop prediction module dataset requires the following columns: State, District, Crop,
Season, Average Temperature, Average Rainfall, Soil Type, Area and Production as these are the major
factors that crops depend on. ‗Production‘ is the dependent variable or the class variable. There are

13
eight independent variables and 1 dependent variable. We achieved this by merging the datasets. The
datasets were merged taking the location as the common attribute in both.

Fig.5.2: Soil and Crop data sample

5.3 Exploratory Data Analysis

It is an approach to analysing datasets to summarize their main characteristics, often with visual
methods. It is also about knowing your data, gaining a certain amount of familiarity with the data, before
one starts to extract insights from it. The idea is to spend less time coding and focus more on the analysis
of data itself. After the data has been collected, it undergoes some processing before being cleaned and
EDA is then performed. After EDA, go back to processing and cleaning of data, i.e., this can be an
iterative process. Subsequently, use the cleaned dataset and knowledge from EDA to perform modelling
and reporting. Exploratory data analysis is generally cross- classified in two ways. First, each method
is either non-graphical or graphical. And second, each method is either univariate or multivariate
(usually just bivariate). It is a good practice to understand the data first and try to gather as many insights
from it. EDA is all about making sense of data in hand. EDA can give us the following:

• Preview data
• Check total number of entries and column types using in built functions. It is a
good practice to know the columns and their corresponding data types.
• Check any null values.
• Check duplicate entries
• Plot distribution of numeric data (univariate and pairwise joint distribution)
• Plot count distribution of categorical data.

14
 Using various built in functions, we can get an insight of the number of values in each column
which can give us information about the null values or the duplicate data. We can also find the mean,
standard deviation, minimum value and the maximum value. This is the basic procedure of EDA. To
get a better insight of the data that is being used, we can plot graphs like the correlation matrix which
is one of the most important concept which gives us a lot of information about how variables (columns)
are related to each other and the impact each of them have on the other. A few other graphs like box
plot and distribution graphs can be plotted too.

Fig.5.3: Correlation Matrix Example

5.4 EDA Performed:

Fig.5.4: EDA Code

Fig.5.5: Output of describe function

15
Fig 5.4 shows the simple code written in python to perform the initial steps in EDA .i.e., finding the
number of columns, the features, the missing values, etc. Fig 5.5 shows the details of each attribute in
tabular form which helps in getting a deeper insight of the attributes.

Fig.5.6: Correlation Matrix of the Proposed System

5.5 Algorithms Used

Machine Learning offers a wide range of algorithms to choose from. These are usually divided
into classification, regression, clustering and association. Classification and regression algorithms come
under supervised learning while clustering and association comes under unsupervised learning.
• Classification: A classification problem is when the output variable is a
category, such as ―red‖ or ―blue‖ or ―disease‖ and ―no disease‖. Example:
Decision Trees
• Regression: A regression problem is when the output variable is a real value,
such as ―dollars‖ or ―weight‖. Example: Linear Regression
• Clustering: A clustering problem is where you want to discover the inherent
groupings in the data, such as grouping customers by purchasing behaviour.
Example: k means clustering
• Association: An association rule learning problem is where you want to
discover rules that describe large portions of your data, such as people that buy
X also tend to buy Y.

16
 Table 5.1: Summary of the Approaches
Algorithm Accuracy

1. Logistic Regression 93.63%

2. Decision Tree 95.45%

3. Random Forest 97.87%

From the Table 5.1, we can conclude that the Random Forest Algorithm gives the best
accuracy for our dataset

5.6 Random Forest

Random forest is a flexible, easy to use machine learning algorithm that produces, even without
hyper-parameter tuning, a great result most of the time. It is also one of the most used algorithms,
because of its simplicity and diversity. It can be used for both classification and regression tasks.
Random forest builds multiple decision trees and merges them together to get a more accurate and stable
prediction. One big advantage of random forest is that it can be used for both classification and
regression problems, which form the majority of current machine learning systems. Another great
quality of the random forest algorithm is that it is very easy to measure the relative importance of each
feature on the prediction. It computes this score automatically for each feature after training and scales
the results so the sum of all importance is equal to one.
The working of Random Forest is as follows:
• Step 1 − First, start with the selection of random samples from a given dataset.
• Step 2 − Next, this algorithm will construct a decision tree for every sample. Then
it will get the prediction result from every decision tree.
• Step 3 − In this step, voting will be performed for every predicted result.
• Step 4 − At last, select the most voted prediction result as the final prediction.

Fig.5.7: Random Forest Flow

17
5.6.1 Another Reason for choosing Random Forest:
The data of a particular crop was taken and passed through two algorithms .i.e.,
Random Forest and another Algorithm that is said to give best results for that crop. The
accuracy achieved in both the algorithms were compared. Rice and Groundnut were chosen
based on the research papers that were found.
• Rice: According to the paper, the best algorithm for rice yield prediction is Linear
Regression . After running both the algorithms, we found a very high difference
between actual value and predicted value in Linear Regression while Random
Forest continued to maintain an accuracy of 90+.
• Groundnut: The Research Paper stated that KNN works best for Groundnut yield
prediction . On running the algorithms, we did not find much difference in the
results in both the algorithms. Hence, we can conclude that Random Forest can be
used as a general algorithm which gives a considerably high accuracy with Good
predictions.

18
 CHAPTER 6
SYSTEM ANALYSIS AND DESIGN

Systems development is a systematic process which includes phases such as planning, analysis,
design, deployment, and maintenance. System Analysis is a process of collecting and interpreting facts,
identifying the problems, and decomposition of a system into its components. System analysis is
conducted for the purpose of studying a system or its parts in order to identify its objectives. It is a
problem solving technique that improves the system and ensures that all the components of the system
work efficiently to accomplish their purpose. Analysis specifies what the system should do. System
Design is a process of planning a new business system or replacing an existing system by defining its
components or modules to satisfy the specific requirements. Before planning, you need to understand
the old system thoroughly and determine how computers can best be used in order to operate efficiently.
System Design focuses on how to accomplish the objective of the system.

6.1 System Architecture

Architecture diagrams can help system designers and developers visualize the high-level,
overall structure of their system or application for the purpose of ensuring the system meets their
users' needs. They can also be used to describe patterns that are used throughout the design. It's
somewhat like a blueprint that can be used as a guide for the convenience of discussing, improving,
and following among a team.

19
 Fig.6.1: System Architecture.

6.2 Flowchart
A flowchart is simply a graphical representation of steps. It shows steps in sequential order
and is widely used in presenting the flow of algorithms, workflow or processes. Typically, a
flowchart shows the steps as boxes of various kinds, and their order by connecting them with
arrows. It originated from computer science as a tool for representing algorithms and programming
logic but had extended to use in all other kinds of processes. Nowadays, flowcharts play an
extremely important role in displaying information and assisting reasoning. They help us visualize
complex processes, or make explicit the structure of problems and tasks. A flowchart can also be
used to define a process or project to be implemented.

20
Fig.6.2: Research overview

21
 CHAPTER 7

SYSTEM REQUIREMENTS SPECIFICATIONS

A software requirements specification (SRS) is a description of a software system to be

developed. It lays out functional and non functional requirements, and may include a set of use cases
that describe user interactions that the software must provide. It is very important in a SRS to list out
the requirements and how to meet them. It helps the team to save upon their time as they are able to
comprehend how are going to go about the project. Doing this also enables the team to find out about
the limitations and risks early on.

A SRS can also be defined as a detailed description of a software system to be developed with
its functional and non-functional requirements. It may include the use cases of how the user is going to
interact with the software system. The software requirement specification document is consistent with
all necessary requirements required for project development. To develop the software system we should
have a clear understanding of Software system. To achieve this we need continuous communication
with customers to gather all requirements.

A good SRS defines how the Software System will interact with all internal modules, hardware,
and communication with other programs and human user interactions with a wide range of real life
scenarios. It is very important that testers must be cleared with every detail specified in this document
in order to avoid faults in test cases and its expected results.

Qualities of SRS

• Correct
• Unambiguous
• Complete
• Consistent
• Ranked for importance and/or stability
• Verifiable
• Modifiable
• Traceable

22
7.1. Basic Requirements

1. Data collection: The dataset used in this project is the data collected from reliable websites and
merged to achieve the desired data set. The sources of our datasets are: https://en.tutiempo.net/ for
weather data and https://www.kaggle.com/srinivas1/agricuture-crops-production-in-india for crop yield
data. It consists of names of the crops, production, area, average temperature, average rainfall (mm),
season, year, name of the states and the districts. ‗Production‘ is the dependent variable or the class
variable. There are eight independent variables and 1 dependent variable.

2. Data Preprocessing: The purpose of preprocessing is to convert raw data into a form that fits
machine learning. Structured and clean data allows a data scientist to get more precise results from an
applied machine learning model. The technique includes data formatting, cleaning, and sampling. Here,
data pre-processing focuses on finding the attributes with null values or invalid values and finding the
relationships between various attributes as well. Data Pre-processing also helps in finding out the impact
of each parameter on the target parameter. To preprocess our datasets we used EDA methodology. All
the invalid and null values were handled by removing that record or giving the default value of that
particular attribute based on its importance.

3. Dataset splitting: A dataset used for machine learning should be partitioned into two subsets —
training and test sets. We split the dataset into two with a split ratio of 80% i.e., in 100 records 80
records were a part of the training set and remaining 20 records were a part of the test set.

4. Model training: After a data scientist has preprocessed the collected data and split it into train and
test can proceed with a model training. This process entails ―feeding‖ the algorithm with training data.
An algorithm will process data and output a model that is able to find a target value (attribute) in new
data an answer you want to get a predictive analysis. The purpose of model training is to develop a
model. We trained our model using the random forest algorithm. On training the model it predicts the
yield on giving the other attributes of the dataset as input.

5. Model evaluation and testing: The goal of this step is to develop the simplest model able to
formulate a target value fast and well enough. A data scientist can achieve this goal through model
tuning. That‘s the optimization of model parameters to achieve an algorithm‘s best performance.

23
7.2. Requirement

1. The access permissions for system data may only be changed by the system‘s data
administrator.
2. Data shall never be viewable at the point of entry or at any other time.
3. Webpage should be responsive to the user Input or to any external interrupt which is of
highest priority and return back to the same state.
4. The webpage must be easy to maintain.
5. It must be user-friendly. Having a user-friendly application is of key importance for the success
of the webpage.

7.2.1. Hardware Requirement

The hardware requirements include the requirements specification of the physical computer
resources for a system to work efficiently. The hardware requirements may serve as the basis for a
contract for the implementation of the system and should therefore be a complete and consistent
specification of the whole system. The Hardware Requirements are listed below:
1. Processor: A processor is an integrated electronic circuit that performs the calculations that run
a computer. A processor performs arithmetical, logical, input/output (I/O) and other basic
instructions that are passed from an operating system (OS). Most other processes are dependent
on the operations of a processor. A minimum 1 GHz processor should be used, although we
would recommend S2GHz or more. A processor includes an arithmetical logic and control unit
(CU), which measures capability in terms of the following:
• Ability to process instructions at a given time
• Maximum number of bits/instructions
• Relative clock speed Learning

Fig.7.1: Processor

The proposed system requires a 2.4 GHz processor or higher.

24
2. Ethernet connection (LAN) OR a wireless adapter (Wi-Fi): Wi-Fi is a family of radio
technologies that is commonly used for the wireless local area networking (WLAN) of devices
which is based around the IEEE 802.11 family of standards. Devices that can use Wi-Fi
technologies include desktops and laptops, smartphones and tablets, TV‘s and printers, digital
audio players, digital cameras, cars and drones. Compatible devices can connect to each other
over Wi- Fi through a wireless access point as well as to connected Ethernet devices and may
use it to access the Internet. Such an access point (or hotspot) has a range of about 20 meters
(66 feet) indoors and a greater range outdoors. Hotspot coverage can be as small as a single
room with walls that block radio waves, or as large as many square kilometres achieved by
using multiple overlapping access points.

Fig.7.2: Wi-Fi

3. Hard Drive: A hard drive is an electro-mechanical data storage device that uses magnetic
storage to store and retrieve digital information using one or more rigid rapidly rotating disks,
commonly known as platters, coated with magnetic material. The platters are paired with
magnetic heads, usually arranged on a moving actuator arm, which reads and writes data to the
platter surfaces. Data is accessed in a random-access manner, meaning that individual blocks
of data can be stored or retrieved in any order and not only sequentially. HDDs are a type of
nonvolatile storage, retaining stored data even when powered off. 32 GB or higher is
recommended for the proposed system.

Fig.7.3: Hard Drive

25
4. Memory (RAM): Random-access memory (RAM) is a form of computer data storage that
stores data and machine code currently being used. A random- access memory device allows
data items to be read or written in almost the same amount of time irrespective of the physical
location of data inside the memory. In today's technology, random-access memory takes the
form of integrated chips. RAM is normally associated with volatile types of memory (such as
DRAM modules), where stored information is lost if power is removed, although non- volatile
RAM has also been developed. A minimum of 2 GB RAM is recommended for the proposed
system.

Fig.7.4: RAM

7.2.2 Software Requirements

The software requirements are description of features and functionalities of the target system.
Requirements convey the expectations of users from the software product. The requirements can
be obvious or hidden, known or unknown, expected or unexpected from client‘s point of view.
1 Jupyter Notebook: The Jupyter Notebook is an open source web application that you can use
to create and share documents that contain live code, equations, visualizations, and text. Jupyter
ships with the IPython kernel, which allows you to write your programs in Python, but there
are currently over 100 other kernels that you can also use. The Jupyter Notebook combines
three components:

• The notebook web application: An interactive web application for writing and running
code interactively and authoring notebook documents.

• Kernels: Separate processes started by the notebook web application that runs users‘ code
in a given language and returns output back to the notebook web application. The kernel
also handles things like computations for interactive widgets, tab completion and
introspection.

• Notebook documents: Self- contained documents that contain a representation of all

content visible in the notebook web application, including inputs and outputs of the

26
 computations, narrative text, equations, images, and rich media representations of objects.
Each notebook document has its own kernel.

Fig.7.5: Jupyter Notebook

2 Python: It is an object-oriented, high-level programming language with integrated dynamic

semantics primarily for web and app development. It is extremely attractive in the field of Rapid
Application Development because it offers dynamic typing and dynamic binding options.
Python is relatively simple, so it's easy to learn since it requires a unique syntax that focuses on
readability. Developers can read and translate Python code much easier than other languages.
In turn, this reduces the cost of program maintenance and development because it allows teams
to work collaboratively without significant language and experience barriers. Additionally,
Python supports the use of modules and a package, which means that programs can be designed
in a modular style and code can be reused across a variety of projects.
3

Fig.7.6: Python

4 Pycharm: PyCharm is the most popular IDE for Python, and includes great features such as
excellent code completion and inspection with advanced debugger and support for web
programming and various frameworks. The intelligent code editor provided by PyCharm
enables programmers to write high quality Python code. The editor enables programmers to
read code easily through colour schemes, insert indents on new lines automatically, pick the
appropriate coding style, and avail context-aware code completion suggestions. At the same
time, the programmers can also use the editor to expand a code block to an expression or logical

27
 block, avail code snippets, format the code base, identify errors and misspellings, detect
duplicate code, and auto-generate code. PyCharm offers some of the best features to its users
and developers in the following aspects
• Code completion and inspection
• Advanced debugging
• Support for web programming and frameworks such as Django and Flask

Fig.7.7: Pycharm

5 Flask: Flask is a lightweight WSGI web application framework. It is designed to make getting
started quick and easy, with the ability to scale up to complex applications. It began as a simple
wrapper around Werkzeug and Jinja and has become one of the most popular Python web
application frameworks. It offers suggestions, but doesn't enforce any dependencies or project
layout. It is up to the developer to choose the tools and libraries they want to use. There are
many extensions provided by the community that make adding new functionality easy.

Fig.7.8: Flask

28
6 Microsoft Visual Studio: Microsoft Visual Studio is an integrated development environment
(IDE) from Microsoft. It is used to develop computer programs, as well as websites, web apps,
web services and mobile apps. Visual Studio uses Microsoft software development platforms
such as Windows API, Windows Forms, Windows Presentation Foundation, Windows Store
and Microsoft Silverlight. It can produce both native code and managed code.

Fig.7.9: Visual Studio

29
 CHAPTER 8
RESULT & DISCUSSION

8.1 Result
In result, firstly we find out the most accuracy algorithm among Logistic Regression, Decision
Tree and Random Forest.
Table 8.1 will show the accuracy levels of the mentioned algorithms.

Table 8.1: Table OF Accuracy Levels

Algorithm Accuracy

1. Logistic Regression 93.63%

2. Decision Tree 95.45%

3. Random Forest 97.87%

Here we found the suitable algorithm is Random Forest with 97.87% accuracy score.
Inputs has to given as shown in the Fig. 8.1

Fig.8.1: Input Screen

The system will predict the output by using Random Forest Algorithm and it display as shown
below. The inputs to be given are Temperature, Humidity, PH, Phosphorus, Nitrogen, Photasium ,
Rainfall and Water Level as shown in fig. 8.1.

30
 Fig.8.2: Output Screen

Hence, the system return the most suitable crop as shown in the fig. 8.2.

31
 CHAPTER 9
CONCLUSION & FUTURE WORK

9.1 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
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 decision in selecting the crop for cultivation such that agricultural sector will be developed by
innovative idea.

9.2 Future Work

The future work will be focused on updating the datasets from time to time to produce accurate
predictions, and the processes can be automated. 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 of the food.
Also, we would be building an application where the farmers can use it as app and converting
the whole system in their regional language.

32
 REFERENCES

[1] Nischitha K, Dhanush Vishwakarma, Mahendra N, Ashwini, Manjuraju M.R “Crop Prediction
using Machine Learning Approaches” 2020.

[2] Bhawana Sharma, Jay Kant Pratap Singh Yadav, Sunita Yadav “Predict Crop Production in India
Using Machine Learning Technique: A Survey” 2020.

[3] Shilpa Mangesh Pande, Dr. Prem Kumar Ramesh, Anmol, B.R Aishwarya, Karuna Rohilla, Kumar
Shourya “Crop Recommender System Using Machine Learning Approach” 2021.

[4] M. Kalimuthu, P. Vaishnavi, M. Kishore “Crop Prediction using Machine Learning” 2020.

[5] Potnuru Sai Nishant, Pinapa Sai Venkat, Bollu Lakshmi Avinash, B. Jabber ”Crop Yield Prediction
based on Indian Agriculture using Machine Learning” 2020.

[6] D. Jayanarayana Reddy, Dr. M Rudra Kumar “Crop Yield Prediction using Machine Learning
Algorithm” 2021.

[7] Namgiri Suresh, N.V.K. Ramesh, Syed Inthiyaz, P. Poorna Priya, Kurra Nagasowmika, Kota.V.N.
Harish Kumar, Mashkoor Shaik, B.N.K. Reddy “Crop Yield Prediction Using Random Forest
Algorithm” 2021.

[8] Fatin Farhan Haque, Ahmed Abdelgawad, Venkata Prasanth Yanambaka, Kumar Yelamarthi “Crop
Yield Analysis Using Machine Learning Algorithms” 2020.

[9] Shivani S. Kale, Preeti S. Patil “A Machine Learning Approach to Predict Crop Yield and Success
Rate” 2020.

[10] Viviliya B, Vaidhehi V “The Design of Hybrid Crop Recommendation System using Machine
Learning Algorithms” 2020.

[11] Pavan Patil, Virendra Panpatil, Prof. Shrikant Kokate “Crop Prediction System using Machine
Learning Algorithms”.

[12] Himani Sharma, Sunil Kumar “A Survey on Decision Tree Algorithms of Classification in Data
Mining”.

[13] Umamaheswari S, Sreeram S, Kritika N, Prasanth DJ, “BIoT: Blockchain-based IoT for
Agriculture”, 11th International Conference on Advanced Computing (ICoAC), 2019 Dec 18 (pp.
324-327). IEEE.

[14] Pijush Samui, Venkata Ravibabu Mandla, Arun Krishna and Tarun Teja “Prediction of Rainfall
Using Support Vector Machine and Relevance Vector Machine”.

33
 APPENDIX

#Creating_Web
import numpy as np
from flask import Flask, request, jsonify, render_template
import pickle

app = Flask(__name__)
model = pickle.load(open('model.pkl', 'rb'))

str=" "

@app.route('/')
def home():
return render_template('index.html')

@app.route('/predict',methods=['POST'])
def predict():
'''For rendering results on HTML GUI '''
int_features = [float(x) for x in request.form.values()]
final_features = [np.array(int_features)]
prediction = model.predict(final_features)

s= str.join(prediction)

# output = round(prediction[0], 2)

return render_template('index.html', prediction_text='The predicted crop is {}'.format(s))

if __name__=="__main__":
app.run(debug=True)

34
#Using_RandomForestAlgorithmInPython
import numpy as np
import pandas as pd

#for data visualization

import matplotlib.pyplot as plt
import pickle
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import train_test_split

df = pd.read_csv('Data.csv')

# print(df.isna().sum())

df.P = df.P.fillna(value = 0)
df.K = df.K.fillna(value = 0)

df.temperature = df.temperature.fillna(value = df.temperature.mean())

df.humidity = df.humidity.fillna(value = df.humidity.mean())
df.ph = df.ph.fillna(value = df.ph.mean())
df.rainfall = df.rainfall.fillna(value = df.rainfall.mean())
df.label = df.label.fillna(value = df.label.mode()[0])

categor_condn = [(df['rainfall'] <= 150),

(df['rainfall'] > 250)]

rating = ['low','high']
df['Water Usage'] = np.select(categor_condn,rating,default = 'medium')

# print(df['Water Usage'].value_counts())
# print(df)
from sklearn.linear_model import LinearRegression as lm

# print(df.isna().sum())

35
label_encoder = LabelEncoder()
df['Water Usage'] = label_encoder.fit_transform(df['Water Usage'])
# df['label'] = label_encoder.fit_transform(df['label'])

# print(df.head())

# # Applying get dummies function on categorical column

x = df.drop('label',axis = 1)
y = df['label']

x_train , x_test , y_train , y_test = train_test_split(x,y, test_size = 0.30 , random_state = 10)

print("Dimension of x_train :",x_train.shape)

print("Dimension of x_test :",x_test.shape)
print("Dimension of y_train :",y_train.shape)
print("Dimension of y_test :",y_test.shape)

from sklearn.ensemble import RandomForestClassifier

# logmodel = lm()
# logmodel.fit(x_train.values,y_train.values)
rfc = RandomForestClassifier(n_estimators = 500,criterion = "entropy")
rfc.fit(x_train.values,y_train.values)
# predict_r = rfc.predict(x_test)

pickle.dump(rfc , open('model.pkl','wb'))

model = pickle.load(open('model.pkl','rb'))
# print(model.predict([[61,38,55,76,52,28,180,2]]))

prediction = rfc.predict((np.array([[61,38,55,76,52,28,180,2]])))
print("The suggested crop is : ",prediction)

prediction = rfc.predict((np.array([[90,42,43,20.87,82,6.5,203,2]])))
print("The suggested crop is : ",prediction)

36