Tellicorp An Ensemble Model To Predict Crop Using Machine Learning Algorithms
Tellicorp An Ensemble Model To Predict Crop Using Machine Learning Algorithms
The goal of this research is to use machine learning techniques to forecast yield in
different districts of India on a yearly basis. The agricultural sector is the prime
occupation of India. Researchers are developing various scientific technologies in
the agriculture field for better yield. In this project, we try to form an ensemble model
using various machine learning algorithms for better rice production. Crop production
prediction utilizing AI Strategies aims to deliver improved outcomes, but the
ensemble model provides better predictive results compared to the individual
algorithm. We tried to use a combination of symmetric machine learning algorithms
to form an ensemble model for better prediction. Here symmetric algorithms such as
random forest, Gradient Boosting, and Logistic Regression are individually used for
the prediction of the yield of rice. While combining all the aforesaid algorithms to
form an ensemble model of a better result. Crop recommendation depends upon soil
characteristics, soil nutrient contents, topography, and other uncontrollable
parameters such as weather conditions, rainfall, humidity, environment temperature.
Modern technologies need to be integrated into farming practices to enhance crop
productivity. The Machine Learning technique is an essential approach for building
an appropriate model for crop recommendation.
iv
TABLE OF CONTENTS
Chapter Title Pg no.
no.
1 INTRODUCTION 1
1.1 OUTLINE OF THE PROJECT 1
1.2 MACHINE LEARNING 1
2 LITERATURE SURVEY 3
3 ML ALGORITHM 6
4 DATASET 8
5 SOFTWARE REQUIREMENTS 14
5.2.1. NumPy 15
5.2.2. Sklenar 15
5.2.3. Pandas 16
5.2.4. Matplotlib 16
5.2.5. Seaborn 17
6 METHODOLOGY 18
6.1 WORKFLOW DIAGRAM 18
v
6.1.1. Random Forest 18
6.1.2 Logistic Regression 20
6.1.3. Gradient Boosting 21
6.1.4. Ensemble Model 23
7 APPLICATION WORKING 25
7.1 Machine Learning 26
7.2. Web Application 26
8 RESULTS AND DISCUSSION 32
9 CONCLUSION 35
REFERENCES 36
APPENDIX 37
A.SOURCE CODE 37
B.SCREENSHOTS 39
C.PLAGIARISM REPORT 42
D. CERTIFICATES 43
E. JOURNAL PAPER 45
F. COPYRIGHT FORM 51
vi
LIST OF FIGURES
4.1 9
Count of data points in the data variable Rainfall, N, P,
K, Temperature, Humidity, and pH
vii
6.10 prediction of ensemble model test and 24
train data
viii
LIST OF TABLES
ix
CHAPTER 1
INTRODUCTION
1
approaches. In ancient computing, algorithms are sets of expressly programmed
directions employed by computers to calculate or downside solve. Machine
learning algorithms instead give computers to coach on knowledge inputs and use
applied math analysis so as to output values that fall inside a particular variable.
Thanks to this, machine learning facilitates computers in building models from
sample knowledge so as to modify decision-making processes supported
knowledge inputs.
2
CHAPTER 2
LITERATURE REVIEW
● The study of [2] (Mythili & Rangaraj, 2021) reveals a new deep learning-based
approach for predicting crop yields. The proposed algorithm PSO-MDNN
recommended a suitable crop recommender model, to recommend a suitable crop
effectively. Its accuracy outperformed different machine learning techniques. The
proposed model predicted yields in untested environments and recommended
crop cultivation. The weight matrices with L2 regularization and PSO are used to
modify the MDNN hyperparameters. The prediction accuracy was enhanced by
using a network structure with optimum weights using PSO.
● [3] (Rajak, Pawar & pendke, 2017) Discuss the ensemble model in conjunction
with the majority voting approach. SVM using ANN as a learner is combined with
Random Tree and NB-classifier to propose a crop with higher efficiency and
accuracy.
● [4] (Kuanr, Rath & Mohanty, 2018) depending on the location and agricultural
preferences, proposes seed, herbicides, and equipment The agricultural yield in
the Kharif season is predicted using fuzzy logic and cosine similarity. Farmers can
predict crop production based on previous yield data and expected weather
conditions.
● The paper [5] (Kumar, Singh, Kumar & Singh, 2015) Crop selection method
(CSM) proposed and advises crop selection to enhance the net production rate.
3
They recommend the sort of crop to be yielded based on parameters such as rain,
temperature, and humidity, as well as the type of soil.
● [7] (Lakshmi, Priya, Sahana & Manjunath, 2018) the goal was to create a crop
growth monitoring recommendation system. The unique qualities utilized were soil
composition, color, drainage, depth, ph. value, water holding capacity, erosion,
permeability, and weather. Map-reduce with KNN suggests quickly and accurately.
● [8] (Akshatha & Shreedhara, 2018) suggested an ensemble model for precision
agriculture that uses different machine learning algorithms to select the crop that is
most suited to the farmer based on the information (factor influencing crop
production) that he provides
● Across the US Corn Belt, homogeneous ensembles produce the most accurate
yield projections. This model could accurately estimate 2019 corn grain yields with
a root mean square error of 866 kg/ha, or 8.5 percent relative root mean square
and could explain almost 77% of the spatiotemporal variation in corn grain
yields.[9](Shahhosseini M et.al 2021)
●The authors [10](Rashid, M., Bari, B. S., Yusup, Y., Kamaruddin, M. A., & Khan,
N.) proposed a system based on a critical evaluation of existing related studies, a
prospective architecture for machine learning-based palm oil yield prediction. This
system will live up to its promise by tackling new research issues in agricultural
yield prediction analysis.
4
●[11](Thomasvan Klompenburga, and AyalewKassahuna) The most used features
are temperature, rainfall, and soil type, and the most applied algorithm is Artificial
Neural Networks in these models. After this observation based on the analysis of
machine learning-based 50 papers, we performed an additional search in
electronic databases to identify deep learning-based studies.
●[12] (Joshua, V.; Priya Harson, S.M.; Kannadasan, R.) did a statistical analysis of
possible MLR approaches and machine learning algorithms such as SVM, GRNN,
RBFNN, and BPNN are examined for evaluation in order to achieve improved
accuracy in agricultural yield prediction. The accuracy level of the various
algorithms is scrutinized using model performance indicate