VR SIDDHARTHA ENGINEERING COLLEGE,

VIJAYAWADA
Department of Computer Science and Engineering

WATER QUALITY PREDICTION AND

CLASSIFICATION USING SVM AND
XGBOOST CLASSIFIERS

Batch No: 13

Batch Members Under the Guidance of

BANAVATH BABY NAGA LEELA (198W1A0571)
Mr. S.BABU M.Tech(ph.D)
CHEEKURIMELLI GANESH KARTHIK(198W1A0576)
YEPURI LAKSHMI NARAYANA (198W1A0592)

1
Abstract
One of the most valuable natural resources ever given to humans is water. The ecosystem and human health
are directly impacted by the water quality. Water is used for many different things, including drinking,
farming, and industrial uses. As a result, predicting and modeling water quality have become essential to
minimizing water pollution. We have used Long Short Term Memory (LSTM) algorithm in Neural Network
for water quality index. In order to achieve a sustainable development, it is crucial to evaluate the
fundamental aspects of the water environment. This research has developed a classification algorithm to
predict the water quality classification (WQC) by designing a web application. The WQC is classified based
on the water quality index (WQI) from 7 parameters in a dataset using Support Vector Machine (SVM) and
Extreme Gradient Boosting (XGBoost). The results from the proposed model can accurately classify the
water quality based on their features. The research outcome demonstrated that the XGBoost model performed
better, with an accuracy of 94%, compared to the SVM model, with only a 67% accuracy. Even better, the
XGBoost resulted in only 6% misclassification error compared to SVM, which had 33%. On top of that,
XGBoost also obtained consistent superior results from validation with an average accuracy of 90%, while
SVM with an average accuracy of 64%. Considering the enhanced performance, XGBoost is concluded to be
better at water quality classification.

Keywords : Water Pollution, Water Quality Index, Water Quality Classification, SVM, XGBoost, LSTM,
Web application, Python , Performance comparison.

2
Presentation Outline
1. Aim and Motivation 11.2. Activity Diagram
2. Title Justification 11.3. Use case Diagram
3. Introduction 11.4. OR any other Diagrams
4.1. (If any) 12. Functional and Non-Functional
4.2. (If any) Requirements
4. Study on Existing Technologies 13. Dataset Description (If any)
5. Gap Analysis 14. Data Preparation (If any)
6. Objectives 15. Timeline Chart
7. Scope 16. Summary
8. Methodology References
9.1. Proposed Model
9.2. Modules of the Proposed Model
9.3 Algorithms
10. SDLC Model
11. UML Diagrams (Any three)
11.1. Sequence Diagram 3
1. Aim and Motivation
Aim:

• To develop an accurate and efficient automated solution for classifying water quality with water quality index
using machine learning.
• To compare the performance of various classifiers and determine the best performing model for detecting the
type of water quality.
• To integrate the best-performing model into a web application with a user-friendly interface.

Motivation:

4
3. Title Justification
 The title clearly states the focus of the research project, which is to analyze the performance of two
machine learning classifiers, SVM and XGBoost for the classification of water quality.
 The title includes important information about the methodology used in the research, which is the
comparison of the performance of the two classifiers.

5
4. Introduction
Instructions
1. A project introduction should include key details about the project that
give the reader enough information to understand the purpose and scope
of the project.
2. Use bullet points or a numbered list to write the content. The sentences
should be neither too short nor too long.
3. In the case of a figure, a caption must be included at the bottom. Multiple
images in a single figure must be labeled as (a)..., (b)..., etc. (If required).
4. Multiple subsections of the introduction section must be labeled as 4.1.,
4.2., etc.

Caption of a figure that

contains multiple images

6
5. Study on Existing Technologies
Title: Title of the article [Reference no.]
Journal Details: Name of the Journals, Vol., Issue, Publisher, Year
Dataset: ORL and Caltech Face
Description:
[Example]
Gorgel and Simsek presented Deep Stacked Denoising Sparse Autoencoders (DSDSA)
for face recognition. It integrated DNN, sparse autoencoders, and smoothing techniques
for robust representation of face samples with different postures, facial expressions, and
occlusions under various lighting conditions. At last, the Random Forest and Multiclass
Support Vector Machine (SVM) classifier trained on the facial features from the ORL
and CalTech face samples had secured 97.50% and 98.21% accuracy, respectively.
Advantages:
1.
2.
Disadvantages:
1.
7
2.
5. Study on Existing Technologies
Instructions
1. You can use bullet points or a numbered list to describe the related work,
or else highlight the major words in the description.
2. Advantages and disadvantages must be specific to your work, i.e., if your
work is application-specific, these must be framed with respect
to application.
3. If your work is algorithm- or model-specific, advantages and
disadvantages must be framed with respect to algorithms and models.
4. Include at least eight recently published journal articles from reputed
publishers. Follow the same templates [slide no. 6] for all the articles.

8
5. Study on Existing Technologies
Table no. : Summary of existing implementations
Algorithms/
S. No. Article Title Journal Details Dataset Advantages Disadvantages
Models
 
1. 1.
 1       2. 2.
3. 3. 

1. 1.
 2     2. 2.
3. 3. 
 

1. 1.
 3     2. 2.
3. 3. 
 
 

1. 1.
4 2. 2.
3. 3. 

Instruction: Present a summary of the eight articles in this table. 9

6. Gap Analysis
Instructions
1. The gap, also considered the missing piece or pieces in the research
literature, is the area that has not yet been explored or is under-
explored.
2. This could be a population or sample (size, type, location, etc.),
research method, data collection and/or analysis, or other research
variables or conditions.
3. Must be numbered. Neither make your texts bold nor italicize them.
4. Try to encapsulate the gap analysis within 5 to 6 points.

10
7. Objectives
Instructions
1. Project objectives are the noteworthy steps needed to complete a
project. 
2. Project objectives must be measurable and contain key performance
indicators that will be used to assess a project's overall success. 
3. Must be numbered. Neither make your texts bold nor italicize them.
4. Your project must contain at least three and a maximum of four
objectives.
5. Objectives should not be broad but specific to your work.
6. Initialize each sentence with the word ‘To’.

11
8. Scope
Instructions
1. Scope is a way to set boundaries on your project and define exactly
what goals, deadlines, and project deliverables you'll be working
towards.
2. Clearly specify at least two scopes of your projects. The sentence
should be neither too short nor too long.
3. Must be numbered. Neither make your texts bold nor italicize them.
4. It must be restricted to current project implementation.

12
9.1. Proposed Model
Instructions
1. The steps of the proposed model resemble a pipeline of sequential and
progressive steps, where each step builds on the input provided by a prior one
and transforms it into information that is suitable for the next step of
processing.
2. It should not be too generic but specific to your work.
3. Figure number and caption must be included at the bottom. Do not describe
the figure in PPT. Use draw.io to plot your proposed model block diagram or
architecture.
4. In the case of deep learning models, you can use the tools for plotting different
network architectures.
https://github.com/ashishpatel26/Tools-to-Design-or-Visualize-Architecture-of-
Neural-Network
Proposed block diagram example Proposed architecture example

13
9.2. Modules of Proposed Model
Module 1 Name:
1.
2.
Module 2 Name:
1.
2.
Module 3 Name:
1.
2.
Module 4 Name:
1.
2.
Instructions
1. Describe the functionality of modules in two to three lines.
2. Must be numbered. Neither make your texts bold nor italicize them.
14
9.3. Algorithms/Pseudo Codes
Instructions
1. Write algorithms, or
pseudocode with inputs and Pseudo Code example
outputs.
2. Use proper mathematical
notations to represent
various mathematical
operations.
3. An algorithm or pseudocode
caption must be numbered
and inserted at the top.
4. To understand pseudo-code
conventions, use this link:
https://onlinelibrary.wiley.co
m/doi/pdf/10.1002/04700297
57.app1

15
10. SDLC Model
Instructions
1. Select the appropriate model for your project execution. To
understand different types of model
https://www.tutorialspoint.com/sdlc/index.htm
2. It should not be generic one, all the phases must be defined with
respect to your project phases.
3. In the case of a machine learning project, select the appropriate
model (pt. 1) to represent the seven stages of the machine learning
life cycle (Figure given bellow).
4. Figure number and caption must be included at the bottom. Do not
describe the figure in PPT.

16
11.1. Sequence Diagram
Instructions
1. Use basic symbols and components in your sequence diagram as per the
requirements in your project. To understand sequence diagram:
https://www.lucidchart.com/pages/uml-sequence-diagram
https://www.youtube.com/watch?v=pCK6prSq8aw
2. Use draw.io or Lucidchart to plot your diagram. Do not include [the brown
color text] in your diagram. These are only for understanding purposes.
3. Figure number and caption must be included at the bottom. Do not describe
the figure in PPT.

Sequence diagram Example

17
11.2. Activity Diagram
Instructions
Activity diagram Example
1. Use activity diagram notations
and symbols properly related
to your project work. To
understand activity diagram:
https://www.lucidchart.com/p
ages/uml-activity-diagram

https://www.youtube.com/wat
ch?v=_tCedK0CfMk

2. Use draw.io or Lucidchart to

plot your diagram. Do not
include [the bold black color
text] in your diagram. These
are only for understanding
purposes.
3. Figure number and caption
must be included at the
18
bottom. Do not describe the
11.3. Use Case Diagram
Instructions
1. Select the correct actors, use cases and subject related to your project, and
represent the relation between them properly. To understand activity
diagram: https://www.lucidchart.com/pages/uml-use-case-diagram
2. Use draw.io or Lucidchart to plot your diagram. Do not include the [brown
color text] in your diagram. These are only for understanding purposes.
3. Figure number and caption must be included at the bottom. Do not
describe the figure in PPT.
A few Use Case diagram examples

19
12. Functional and Non-Functional Requirements
1. Functional requirements:  Specifically demands as basic facilities that
the system should offer. All these functionalities need to be necessarily
incorporated into the system.
Refer this video to frame your functional requirements:
https://www.youtube.com/watch?v=D6R7njeK-qg
2. Non-Functional Requirements: The quality constraints that the system
must satisfy according to the project.
Refer the following video to frame your non-functional
requirements:
https://www.youtube.com/watch?v=2DwpAyzJrVQ
To understand functional and non-functional requirements:
https://www.geeksforgeeks.org/functional-vs-non-functional-requirements/
Instructions:
3. Present the functional and non-functional requirements in tabular
format.
4. The table caption must be at the top of the table. Do not Describe. 20
13. Dataset Description [Example]
Name of the Dataset: UNIMIB 2016 [Reference no.]
Description: This database can be used for food recognition, segmentation as well
as quantity estimation. It includes multiple food item images. This dataset was
captured from The canteen situated within the ‘University of Milano-Bicocca’
campus. The particularities of this setting are that each image depicts different foods
on a tray, and some foods (e.g., fruit, bread and dessert) are placed on the place
mats rather than on plates. The images present visual distortions as well as
illumination changes due to shadows.

Classes: 73
Number of Images : 1024
Train set size: 650 (Predefined, else specify the __%)
Test set size: 374 (Predefined, else specify the __%)
Validation set size: ___ (If any)
Image resolution: 480 x 360 pixels

Instructions:
1. You can use bullet points or a numbered list to describe the dataset, else
highlight the major words in the description.
2. If you are using multiple datasets, describe those in separate slides. 21
14. Data Preparation
Instructions:
1. Explain the data preparation procedures, algorithms, or pseudocode
with inputs and outputs. (Vary depending on the chosen domain and
your work) OR block diagram/architecture of data preparation
technique. (If required)
2. Display the input and output with proper captions (If required). All the
figures must be centered.
3. In the case of a figure, a caption must be included at the bottom.
Multiple images in a single figure must be labeled as (a)..., (b)..., etc.
(If required)
To understand pseudo-code conventions, use this link:
https://onlinelibrary.wiley.com/doi/pdf/10.1002/0470029757.app1

Caption of a figure that

contains multiple images
22
15. Timeline Chart
Table no. : Timeline chart for Mini Project – I work

Instruction: The content of this table can be modified based on your project work. If
actions 7, 8, and 9 are part of your project, those must be reflected in your UML
diagrams, functional and non-functional requirements. 23
16. Summary
Here, we have discussed
1. The real-world problem that motivate us to carry out this project.
2.
3.
4.
…

Instructions
1. Must be numbered, and do not italicize your text.
2. Try to encapsulate the summary of your presentation within 6 to 7
points.

24
References
1. Vimala, C., & Priya, P. A. (2019). Artificial neural network-based wavelet
transform technique for image quality enhancement. Computers & Electrical
Engineering, 76, 258-267.

Instructions
2. The same order must be maintained throughout the PPT.
3. In the case of online materials, unpublished work, or online videos, use the
following URL to generate APA references.
  https://www.scribbr.com/citation/generator/apa/

25
 Insert Slide Numbers

1. Slide numbers are important.

2. It not only helps reviewers ask any questions
but also helps you answer them.

[Don’t include this slide to your PPT]

26