Visvesvaraya Technological University

Jnana Sangama, Belagavi-590010

Mini Project Synopsis

ON
“ IOT Based Air Quality Index Monitoring Using ESP32”
Submitted in partial fulfillment for the requirements for the sixth semester
BACHELOR OF ENGINEERING
IN
ELECTRICAL AND ELECTRONICS ENGINEERING
For the Academic Year 2021-2022
Submitted by:

AKSHATHA K 1MV20EE400
CHAITRA C 1MV20EE403
SUNIL R
1MV20EE407

Under the guidance of :

Dr.NANDA SHIVAMUGGI
Asst. Prof & Guide Dept. Of EEE,
Sir M Visvesvaraya Institute of Technology, Bangalore-562157

DEPARTMENT OF ELECTRICAL AND ELECTRONICS

ENGINEERING
SIR M. VISVESVARAYA INSTITUTE OF TECHNOLOGY
Hunasamaranahalli, Bangalore – 562157
 INDEX

SL.No Title Page No.

1 ABSTRACT 3

2 LITERATURE SURVEY 4

3 PROBLEM DEFINITION AND OBJECTIVE 5

4 INTRODUCTION 6

5 REQUIREMENT SPECIFICATION 7

6 DESIGN AND METHODOLOGY 8

Page | 2
 ABSTRACT

Loans are the core business of banks. The main profit comes directly from the loan’s interest.
A dataset containing numerous rows and columns of specific details of customer id, name,
gender, age, income, profession, credit score, location, interest rate, loan amount and loan
sanctioned. This project involves analysis of the given dataset and utilize some Machine
Learning Algorithms to predict the loan amount.

In this project, we’ll build a predictive model to predict the loan amount to be sanctioned to an
applicant using the other data sets. We will prepare the data using Jupyter Notebook and use
various models to predict the target variable.

Respective performances of various algorithms are compared to search out one that most
closely fits the available data set. the ultimate prediction model is evaluated against test data
and accuracy is obtained. We draw conclusions from the obtained plots and models, and the
available data list to predict the final loan amount.

To do this project we will use the concept of Machine learning using regression approach to
compute the final possible loan amount sanctioned and R2 score.

Page | 3
 LITERATURE SURVEY

Numerous pieces of literature about loan prediction have been published already and are
available for public usage.
1. Paper Name: Loan Prediction by using Machine Learning Models Authors: Pidikiti
Supriya, Myneedi Pavani, Nagarapu Saisushma.
Description: Data collection and pre-processing, applying machine learning models, training,
and testing the data were the modules covered in this paper. Outlier detection and removal, as
well as imputation removal processing, were done during the pre-processing stage. To predict
the chances of current status regarding the loan approval process, SVM, DT, KNN, and
gradient boosting models were used in this method. To divide the dataset into training and
testing processes, the 80:20 rule was used. Experimentation concluded that the Decision Tree
has significantly higher loan prediction accuracy than the other models.
Results: Accuracy achieved: 0.811.
Model used: Decision Tree.

2. Paper Name: Credit Risk Analysis and Prediction Modelling of Bank Loans Using R
Authors: Sudhamathy G. Description: Using the R package, this paper proposed a risk analysis
method for sanctioning a loan for customers. Data selection, pre-processing, feature extraction
and selection, building the model, prediction, and evaluation were among the steps involved
in developing the model. The USI repository provided the dataset for evaluation and
prediction. Because the most important and time-consuming step is pre-processing,
classification and clustering techniques in R were used to prepare the data for further use. The
decision tree classifier was then built using the preprocessed dataset.
Results: Precision achieved: 0.833.

Page | 4
 PROBLEM DEFINITION AND OBJECTIVE

Problem Definition
Customer initially applies for loan, then banks validates the customers eligibility for loan. The
bank then processes the loan amount to be sanctioned.

• Predict the loan amount to be sanctioned once the parameters are given consequently.
• Measure the R2 Score of the Machine Learning Algorithm used to predict the data.
• Use Exploratory Data Analysis and plot graphs to illustrate the relationship between
various parameters involved in that graph.
• Compare the R2scores of all the models and deduce which model has the highest R2
score.

Objective

• Identify the type of model to use in the problem statement. Regression in this case
because we have to predict the loan amount to be sanctioned.
• Clean the given dataset for any unwanted values and pass the dataset through a specific
Machine Learning Algorithm.
• Fit the data accordingly and plot the graph between the tested data and predicted output
and identify the accuracy score of the particular Machine Learning Algorithm.
• Repeat the prices with a few more Machine Learning Algorithms and note down the
corresponding accuracy value of each model.
• Plot the graphs to demonstrate Exploratory Data Analysis and explain the relationship
between various parameters depicted in the specific graphs. A minimum of 7 graphs are
to be plotted.

Page | 5
 INTRODUCTION

Python and Machine Learning Internship are offered in cooperation between two companies,
TakeiteasyEngineers and IC Solutions. It initially consisted of a week of training followed by
one-month internship that must be completed in order to receive the internship certificate.

The course began with the presentation from our instructor Abhishek sir with the basics of
Python such as mathematical operations and functions for machine learning course and precise
syntax writing. Two major Python libraries were introduced, namely NumPy and Pandas,
which are used for data generation, data visualization, manipulation and cleaning. Then we
came across another Python library called Matplotlib, which is mainly used to plot and
visualize data in an imagery way and to draw adequate conclusions from the parameters
provided and used in the problem statement and the set of data.

In finance, a loan lending of money by an individual, organizations or banks. The borrower or

recipient enters into a debt and is generally up for paying the interest on that debt until the
principal amount borrowed is repaid.

We created a predictive model to predict the amount of loan that will be sanctioned to an
applicant using the other records by machine learning techniques such as the regression
approach to calculate the final possible sanctioned loan amount and the R2 score.

Page | 6
 REQUIREMENT SPECIFICATION

Hardware Requirements

• Storage – 500GB.
• Ram – 4GB DDR4.
• Processor – 10th Generation Intel® Core™ i5-1035G1 Processor.
• CPU- 2 x 64-bit 2.8 GHz.

Software Requirements

• Python Compiler to write codes.

• NumPy, Pandas, Seaborn, Matplotlib.
• Anaconda Software and Jupyter Notebook installed from it.
• Data Sets for the prediction of loan amount.

Page | 7
 DESIGN AND METHODOLOGY

Since the problem of predicting the approval of a loan application is a classification problem,
the model is trained using classification algorithms like Logistic Regression, Decision Tree,
Random Forest Classifier, Support Vector Machine. The steps involved in building the model
are :-

Page | 8