Department of

GALGOTIAS Information Technology

COLLEGE OF ENGINEERING & TECHNOLOGY

A FINAL YEAR PROJECT PRESENTATION

ON

PRECISION AGRICULTURE USING MACHINE LEARNING

BY GROUP NO.: 23IT717

Purushottam Varshney (2000970130087)

Paras Chandra (2000970130074)
Rohan Kumar Sinha (2000970130095)

Under the supervision of

Ms. Sakshi Malhotra

2023-24

1
 Department of Information Technology

OUTLINE
Problem Statement
Introduction
Literature Survey
Methodology
Description of Modules
Implementation Details
Team Functioning

2
 Department of Information Technology

| Problem Statement
• To demonstrate the application of Machine Learning in the field of precision agriculture for
improved resource use efficiency, productivity, quality, profitability and sustainability of
agricultural production.

• To develop an application that can recommend proper crops to be sown based on the condition
of soil parameters and climate and also to provide fertilizer related recommendations required
for the crop.

3
 Department of Information Technology

| Introduction
• Smart agriculture, also known as precision agriculture, is an innovative approach to farming that
integrates technology and data analytics to optimize crop yields, reduce waste, and minimize
environmental impact.

• In India, the crop yield is not up to the expectation. This happens due to low soil fertility, not growing
desired crops according to the need of the soil, exhaustive over use of fertilizer.

• The present work focuses on recommendation of suitable crops to be sown in the particular field or
soil.

• Our work also facilitates endorsing farmer in sprinkling the required amount of fertilizer in their field
for the particular crop by using the N, P, K values.

• The proposed system will give suggestion to the farmer to meet up the required deficient nutrient and
minerals in the soil.

4
 Department of Information Technology

| Literature Survey

Problem
Paper Methodology Dataset Outcomes
Statement
Efficient Crop Yield Recommendation
Crop recommendation using NPK quantity data collection, Crop and Regional Crop prediction and quantity
System Using Machine Learning for
soil parameters fertilizer recommendation, SVM dataset of fertilizer
Digital Farming, Volume 10, No.1, 2021
A Review of Wireless Sensors and Reviewed applications of Concluded the best
Use cases of technologies in different areas
Networks’ Applications in Agriculture, wireless sensors and WSN NA technologies that are usable
like irrigation, fertilization, etc.
Volume 36, Issue 2, 2014 technologies. in cost effective manner.

Crop Recommendation System using Proposed system of IoT and Produced informed decisions
IoT, Convolutional Neural Network, KNN,
Machine Learning, Volume 7, Issue 3 ML is enabled for soil Kaggle about which crop to
Naïve bayes, Regression
May-June 2021 testing using sensors. plant based on soil condition.

Applications of Smartphone-Based Reviewed the papers and Use cases of different phone sensors like Camera and GPS were found
Sensors in Agriculture: A Systematic smartphone-based camera, GPS, accelerometer, microphone, NA to be the dominant and most
Review of Research, Volume 2015 applications. etc., are discussed. usable sensors in the study.

IOT Based Smart Agriculture Monitoring To develop an IoT based

Used various sensors, Wi-Fi and camera Data collected Able to monitor the crop and
System, Volume: 5 Issue: 2, February system to monitor the
with microcontrollers to monitor fields. through sensors. field remotely.
2017 various crop aspects.

5
 Department of Information Technology

| Methodology
• Our application development is divided into 4 modules-
1. Dataset preparation module
2. Backend module
3. Frontend module
4. Integration module

• Dataset Preparation Module –

- Collected from previous research works and Kaggle dataset repository.
- Data cleaning and modification.
- Final dataset preparation.

• Backend Module –
- Various ML algorithms’ accuracy comparison.
- Model training with highest accuracy algorithm.

6
 Department of Information Technology

| Methodology
• Frontend Module –
- Application UI/UX designing and implementation.
- Used React JS, React-Router-Dom, Tailwind CSS and other libraries.
- Wiring with functionalities.

• Integration Module –
- Synchronization of backend and frontend.
- Used Flask to connect frontend with backend.

7
 Department of Information Technology

| Description of Modules
• Backend Module –
- It takes the prepared dataset and splits it into train and test sets.

- The training set is fed into Naïve Bayes algorithm after

comparison of accuracy.

- The model then learns and predicts the crop

on test data.

- The model also suggests fertilizer based on

crop and available soil nutrients.

8
 Department of Information Technology

| Description of Modules
• Frontend Module –
- User selects the recommendation field and enters values of parameters required.

- The parameters are then send to integration module via API call.

- The integration module utilizes the backend model and makes recommendations.

- Those recommendation are then

displayed to the user on the UI.

9
 Department of Information Technology

| Implementation

Algorithm Accuracy
Naïve Bayes 96.13%

Decision Tree 81.27

SVM 34.18%

Logistic Regression 87.72%

Random Forest 95.22%

Table.1: Comparison of accuracy of various algorithms over the dataset.

10
 Department of Information Technology

| Implementation

Fig.1: Fertilizer recommendations made at the backend.

11
 Department of Information Technology

| Implementation

Fig.2: Recommendations at the frontend.

12
 Department of Information Technology

| Team Functioning

Member’s Name Tasks

Purushottam Varshney Backend (Model selection & model training)

Paras Chandra Frontend + System integration

Rohan Kumar Sinha Dataset Preparation and Research work

13
Thank You