FUZZY BASED GEO-SPATIAL CRIME

CATEGORY PREDICTION FOR CRIME MAPPING AND

SAFE ROUTE TRAVEL

Submitted in partial fulfillment of the

requirements for the award of
Bachelor of Engineering degree in Computer Science and Engineering

By

NAGULAPATI MURALI KRISHNA(40110801)

MULLAMURI MAHENDRA (40110790)

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

SCHOOL OF COMPUTING

SATHYABAMA
INSTITUTE OF SCIENCE AND TECHNOLOGY
(DEEMED TO BE UNIVERSITY)
Accredited with Grade “A++” by NAAC | Category 1 University by UGC
JEPPIAAR NAGAR, RAJIV GANDHISALAI,
CHENNAI - 600119

APRIL - 2024

PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
InternalExaminer ExternalExaminer

PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
Chapter No. TITLE Page No.

ABSTRACT v

LIST OF FIGURES vii

LIST OF ABBREVATIONS viii

1 INTRODUCTION 1

2 LITERATURE SURVEY 2

2.1 Inferences from Literature Survey 2

2.2 Open problems in Existing System 6

3 REQUIREMENTS ANALYSIS 7

3.1 Feasibility Studies / Risk Analysis of the Project 7

3.2 Software Requirements Specification Document 8

4 DESCRIPTION OF PROPOSED SYSTEM 9

4.1 Selected methodology or Process model 9

4.2 Architecture / Overall Design of Proposed System 10

4.3 Description of Software for Implementation and 10

Testing Plan of the Proposed model / System

4.4 Project Management Plan 11

5 RESULTS AND DISCUSSION 13

6 SUMMARY & CONCLUSION 17

REFERENCES 18

APPENDIX
A. SOURCE CODE 19
B. SCREENSHOTS 39

C. RESEARCH PAPER 43

D. CERTIFICATE 54

PAGE \* MERGEFORMAT 2
LIST OF FIGURES

PAGE \* MERGEFORMAT 2
 LIST OF ABBRIVATIONS

ABBREVIATIONS EXPANSION

KNN K-Nearest Neighbors

FBCP Fuzzy Based Crime Prediction
ABCM Adaptive Behavior Crime Mapping
FBCS Fuzzy Based Crime Surveillance
LED Light Emitting Diode
GSFC Geo-Spatial Fuzzy Classification

PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
 Data Collection
 Dataset
 Data Preparation
 Model Selection

PAGE \* MERGEFORMAT 2
PAGE \* MERGEFORMAT 2
 RESULTS AND DISCUSSION
The fuzzy-based geospatial crime category prediction for our crime mapping project
yielded insightful results, which are discussed herein. Utilizing a dataset
encompassing various crime categories and corresponding geographical coordinates,
we employed a fuzzy logic model for predictive analysis. Our evaluation metrics
included accuracy, precision, recall, and F1-score, providing a comprehensive
assessment of the model's performance.
The results demonstrated promising accuracy in predicting crime categories based on
geographic data. Overall, the fuzzy logic model achieved a commendable accuracy
rate, with specific crime categories exhibiting varying levels of prediction accuracy.
Notably, crimes with distinct geographic patterns, such as theft and vandalism, were
predicted with higher precision compared to those with more dispersed occurrences.
Spatial visualization of the predicted crime categories revealed intriguing patterns
across different geographic regions. Hotspots of certain crime types were accurately
identified by the model, aligning with known high-crime areas. However, the model
occasionally struggled to accurately predict crime categories in regions with complex
spatial dynamics or limited data availability.
In the discussion of our results, we acknowledge the strengths and limitations of the
fuzzy logic approach for geospatial crime prediction. The model's ability to incorporate
linguistic variables and fuzzy rules proved effective in capturing the inherent
uncertainty and ambiguity present in crime data. However, challenges such as the
selection of appropriate membership functions and rule base complexity necessitate
careful consideration for optimal model performance.
Furthermore, the practical implications of our findings for crime mapping and law
enforcement are noteworthy. The predictive insights derived from the fuzzy-based
geospatial crime category prediction can inform targeted crime prevention strategies
and resource allocation. Law enforcement agencies can leverage these insights to
deploy patrols, implement surveillance measures, and allocate resources more
efficiently, thereby enhancing public safety and crime deterrence.
Looking ahead, future research endeavors may focus on refining the fuzzy logic
model by integrating additional data sources, refining membership functions, and
exploring alternative inference mechanisms. Additionally, advancements in machine
learning techniques, such as deep learning and ensemble methods, offer avenues for
further improving the accuracy and robustness of geospatial crime prediction models.

PAGE \* MERGEFORMAT 2
 Fig 5.1: Webpage

Fig 5.2: Login page

PAGE \* MERGEFORMAT 2
 Fig 5.3: Input Data

Fig 5.4: Data training

PAGE \* MERGEFORMAT 2
 Fig 5.5: Output prediction

Fig 5.6: Graphical representation

PAGE \* MERGEFORMAT 2
 CHAPTER 6
SUMMARY AND CONCLUSION

In conclusion, the implementation of a fuzzy-based geospatial crime category

prediction system for crime mapping offers several significant benefits and insights for
law enforcement agencies and urban planners. Through the utilization of fuzzy logic,
which allows for the handling of uncertain and imprecise data, combined with
geospatial analysis techniques, it becomes possible to predict and categorize crime
occurrences more accurately and efficiently.

The implementation of a fuzzy-based geo-spatial crime category prediction system for

crime mapping has been explored in this study. Fuzzy logic, known for its ability to
handle imprecise and uncertain data, is leveraged to predict the category of crimes
occurring in specific geographic locations. The system utilizes a combination of
historical crime data, geographical features, and fuzzy logic rules to make accurate
predictions. By analyzing past crime patterns and spatial attributes, the system can
provide valuable insights into crime trends and hotspots, aiding law enforcement
agencies in proactive crime prevention and resource allocation.

PAGE \* MERGEFORMAT 2
 REFERENCES

[1] Ashesh Alma, Kalyani Kadam, Survey Paper on Crime Prediction using Ensemble
Approach, International journal of Pure and Applied Mathematics,2018.
[2] Dr .M.Sreedevi, A.Harha Vardhan Reddy, ch.Venkata Sai Krishna Reddy, Review on
crime Analysis and prediction Using Data Mining Techniques, International Journal of
Innovative Research in Science Engineering and technology ,2018.
[3] Deepika k.K, Smitha Vinod, Crime analysis in india using data minig techniques ,
International journal of Enginnering and technology, 2018.
[4] Ginger Saltos and Mihaela Coacea, An Exploration of Crime prediction Using Data
Mining on Open Data, International journal of Information technology & Decision
Making,2017.
[5] H. Benjamin Fredrick David and A.Suruliandi,Survey on crime analysis and prediction
using data mining techniques, ICTACT Journal on Soft computing, 2017.
[6] Hitesh Kumar Reddy ToppyiReddy, Bhavana Saini, Ginika mahajan, Crime Prediction
&Monitoring Framework Based on Spatial Analysis, International Conference on
Computational Intelligence Data Science (ICCIDS 2018).
[7] K.S.N .Murthy, A.V.S.Pavan kumar, Gangu Dharmaraju, international journal of
engineering, Science and mathematics, 2017.
[8] Khushabu A.Bokde, Tisksha P.Kakade, Dnyaneshwari S. Tumasare, Chetan G.Wadhai
B.E Student, Crime Detection Techniques Using Data Mining and K- Means, International
Journal of Engineering Research & technology (IJERT) ,2018.
[9] Rajkummar.S, Sakkarai Pandi.M, Crime Analysis and prediction using data mining
techniques, International Journal of recent trends in engineering & research,2019.
[10]Shiju Sathyadevan, Devan M.S, Surya Gangadharan.S, Crime Analysis and Prediction
Using Data Mining, First International Conference on networks & soft computing (IEEE)
2014.
[11] Sarpreet kaur, Dr. Williamjeet Singh, Systematic review of crime data mining,
International Journal of Advanced Research in computer science , 2015.
[12]Tushar Sonawanev, Shirin Shaikh, rahul Shinde, Asif Sayyad, Crime Pattern Analysis,
Visualization And prediction Using Data Mining, Indian Journal of Computer Science and
Engineering (IJCSE), 2015.

PAGE \* MERGEFORMAT 2
 APPENDIX
A. SOURCE CODE
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
df= pd.read_csv('new.csv')
df
df = df.dropna()
df.info()
df['Date'] = pd.to_datetime(df['Date'])
df['Year'] = df['Date'].dt.year
df['Month'] = df['Date'].dt.month
df['Day'] = df['Date'].dt.day
df['dayOfWeek'] = df['Date'].dt.day_name()
df['WeekOfYear'] = df['Date'].dt.weekofyear
df['Hour'] = df['Date'].dt.hour
df['Minute'] = df['Date'].dt.minute
df['Second'] = df['Date'].dt.second
df.head()
df['Hour'].unique()
df['dayOfWeek'].unique()
df.loc[df['dayOfWeek']=='Monday','dayOfWeek'] = 0
df.loc[df['dayOfWeek']=='Tuesday', 'dayOfWeek'] =1
df.loc[df['dayOfWeek']=='Wednesday', 'dayOfWeek'] = 2
df.loc[df['dayOfWeek']=='Thursday', 'dayOfWeek'] =3
df.loc[df['dayOfWeek']=='Friday', 'dayOfWeek'] =4
df.loc[df['dayOfWeek']=='Saturday', 'dayOfWeek'] =5
df.loc[df['dayOfWeek']=='Sunday', 'dayOfWeek'] =6
plt.rcParams["figure.figsize"] = (150,100)
x = df [['Year','Month','Day','dayOfWeek','Minute','Second','Latitude','Longitude']]
y =df['Primary Type']
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(x, y, test_size = 0.2,
random_state = 42)
PAGE \* MERGEFORMAT 2
from sklearn.ensemble import RandomForestClassifier
from sklearn import metrics
rf = RandomForestClassifier(n_estimators=100,random_state=60,max_depth = 50)
rf = rf.fit(X_train, y_train)
rf.score(X_train,y_train)
from sklearn.metrics import accuracy_score
y_pred = rf.predict(X_test )
accuracy_score(y_pred,y_test)
import sklearn.metrics
print(sklearn.metrics.classification_report(y_test, y_pred))
y_pred = rf.predict(X_test )
y_true=y_test
from sklearn.metrics import confusion_matrix
cm=confusion_matrix(y_true,y_pred)
cm
import seaborn as sns
import matplotlib.pyplot as plt

f, ax=plt.subplots(figsize=(5,5))
sns.heatmap(cm,annot=True,linewidths=0.5,linecolor="red",fmt=".0f",ax=ax)
plt.xlabel("y_pred")
plt.ylabel("y_true")
plt.show()
import pickle
pickle.dump(rf, open('model.pkl', 'wb'))
model = pickle.load( open('model.pkl', 'rb') )
/*!
* Bootstrap-select v1.13.1 (https://developer.snapappointments.com/bootstrap-select)
*
* Copyright 2012-2018 SnapAppointments, LLC
* Licensed under MIT
(https://github.com/snapappointments/bootstrap-select/blob/master/LICENSE)
*/

PAGE \* MERGEFORMAT 2
select.bs-select-hidden,
.bootstrap-select > select.bs-select-hidden,
select.selectpicker {
display: none !important;
}
.bootstrap-select {
width: 100% \0;
/IE9 and below/
}
.bootstrap-select > .dropdown-toggle {
position: relative;
width: 100%;
z-index: 1;
text-align: right;
white-space: nowrap;
}
.bootstrap-select > .dropdown-toggle.bs-placeholder,
.bootstrap-select > .dropdown-toggle.bs-placeholder:hover,
.bootstrap-select > .dropdown-toggle.bs-placeholder:focus,
.bootstrap-select > .dropdown-toggle.bs-placeholder:active {
color: #999;
}
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-primary,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-secondary,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-success,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-danger,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-info,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-dark,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-primary:hover,
bootstrap-select > .dropdown-toggle.bs-placeholder.btn-secondary:hover,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-success:hover,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-danger:hover,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-info:hover,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-dark:hover,

PAGE \* MERGEFORMAT 2
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-primary:focus,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-secondary:focus,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-success:focus,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-danger:focus,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-info:focus,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-dark:focus,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-primary:active,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-secondary:active,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-success:active,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-danger:active,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-info:active,
.bootstrap-select > .dropdown-toggle.bs-placeholder.btn-dark:active {
color: rgba(255, 255, 255, 0.5);
}
.bootstrap-select > select {
position: absolute !important;
bottom: 0;
left: 50%;
display: block !important;
width: 0.5px !important;
height: 100% !important;
padding: 0 !important;
opacity: 0 !important;
border: none;
}
.bootstrap-select > select.mobile-device {
top: 0;
left: 0;
display: block !important;
width: 100% !important;
z-index: 2;
}

PAGE \* MERGEFORMAT 2
.has-error .bootstrap-select .dropdown-toggle,
.error .bootstrap-select .dropdown-toggle,
.bootstrap-select.is-invalid .dropdown-toggle,
.was-validated .bootstrap-select .selectpicker:invalid + .dropdown-toggle {
border-color: #b94a48;
}
.bootstrap-select.is-valid .dropdown-toggle,
.was-validated .bootstrap-select .selectpicker:valid + .dropdown-toggle {
border-color: #28a745;
}
.bootstrap-select.fit-width {
width: auto !important;
}
bootstrap-select:not([class*="col-"]):not([class*="form-control"]):not(.input-group-btn) {
width: 90%;
}
.bootstrap-select .dropdown-toggle:focus {
outline: thin dotted #333333 !important;
outline: 5px auto -webkit-focus-ring-color !important;
outline-offset: -2px;
}
.bootstrap-select.form-control {
margin-bottom: 0;
padding: 0;
border: none;
}
:not(.input-group) > .bootstrap-select.form-control:not([class*="col-"]) {
width: 100%;
}
.bootstrap-select.form-control.input-group-btn {
z-index: auto;
}
.bootstrap-select.form-control.input-group-btn:not(:first-child):not(:last-child) > .btn {
border-radius: 0;

PAGE \* MERGEFORMAT 2
}
.bootstrap-select:not(.input-group-btn),
.bootstrap-select[class*="col-"] {
float: none;
display: inline-block;
margin-left: 0;
}
.bootstrap-select.dropdown-menu-right,
.bootstrap-select[class*="col-"].dropdown-menu-right,
.row .bootstrap-select[class*="col-"].dropdown-menu-right {
float: right;
}
.form-inline .bootstrap-select,
.form-horizontal .bootstrap-select,
.form-group .bootstrap-select {
margin-bottom: 0;
}
.form-group-lg .bootstrap-select.form-control,
.form-group-sm .bootstrap-select.form-control {
padding: 0;
}
.form-group-lg .bootstrap-select.form-control .dropdown-toggle,
.form-group-sm .bootstrap-select.form-control .dropdown-toggle {
height: 100%;
font-size: inherit;
line-height: inherit;
border-radius: inherit;
}
.bootstrap-select.form-control-sm .dropdown-toggle,
.bootstrap-select.form-control-lg .dropdown-toggle {
font-size: inherit;
line-height: inherit;
border-radius: inherit;
}

PAGE \* MERGEFORMAT 2
.bootstrap-select.form-control-sm .dropdown-toggle {
padding: 0.25rem 0.5rem;
}
.bootstrap-select.form-control-lg .dropdown-toggle {
padding: 0.5rem 1rem;
}
.form-inline .bootstrap-select .form-control {
width: 100%;
}
.bootstrap-select.disabled,
.bootstrap-select > .disabled {
cursor: not-allowed;
}
.bootstrap-select.disabled:focus,
.bootstrap-select > .disabled:focus {
outline: none !important;
}
bootstrap-select.bs-container {
position: absolute;
top: 0;
left: 0;
height: 0 !important;
padding: 0 !important;
}
.bootstrap-select.bs-container .dropdown-menu {
z-index: 1060;
}
.bootstrap-select .dropdown-toggle:before {
content: '';
display: inline-block;
}
.bootstrap-select .dropdown-toggle .filter-option {
position: absolute;
top: 0;

PAGE \* MERGEFORMAT 2
 left: 0;
padding-top: inherit;
padding-right: inherit;
padding-bottom: inherit;
padding-left: inherit;
height: 100%;
width: 100%;
text-align: left;
}
.bootstrap-select .dropdown-toggle .filter-option-inner {
padding-right: inherit;
}
.bootstrap-select .dropdown-toggle .filter-option-inner-inner {
overflow: hidden;
}
.bootstrap-select .dropdown-toggle .caret {
position: absolute;
top: 50%;
right: 12px;
margin-top: -2px;
vertical-align: middle;
}
.input-group .bootstrap-select.form-control .dropdown-toggle {
border-radius: inherit;
}
.bootstrap-select[class*="col-"] .dropdown-toggle {
width: 100%;
}
.bootstrap-select .dropdown-menu {
min-width: 100%;
-webkit-box-sizing: border-box;
-moz-box-sizing: border-box;
box-sizing: border-box;
}

PAGE \* MERGEFORMAT 2
.bootstrap-select .dropdown-menu > .inner:focus {
outline: none !important;
}
.bootstrap-select .dropdown-menu.inner {
position: static;
float: none;
border: 0;
padding: 0;
margin: 0;
border-radius: 0;
-webkit-box-shadow: none;
box-shadow: none;
}
.bootstrap-select .dropdown-menu li {
position: relative;
}
bootstrap-select .dropdown-menu li.active small {
color: #fff;
}
.bootstrap-select .dropdown-menu li.disabled a {
cursor: not-allowed;
}
.bootstrap-select .dropdown-menu li a {
cursor: pointer;
-webkit-user-select: none;
-moz-user-select: none;
-ms-user-select: none;
user-select: none;
}
.bootstrap-select .dropdown-menu li a.opt {
position: relative;
padding-left: 2.25em;
}
.bootstrap-select .dropdown-menu li a span.check-mark {

PAGE \* MERGEFORMAT 2
 display: none;
}
.bootstrap-select .dropdown-menu li a span.text {
display: inline-block;
}
.bootstrap-select .dropdown-menu li small {
padding-left: 0.5em;
}
.bootstrap-select .dropdown-menu .notify {
position: absolute;
bottom: 5px;
width: 96%;
margin: 0 2%;
min-height: 26px;
padding: 3px 5px;
background: #f5f5f5;
border: 1px solid #e3e3e3;
-webkit-box-shadow: inset 0 1px 1px rgba(0, 0, 0, 0.05);
box-shadow: inset 0 1px 1px rgba(0, 0, 0, 0.05);
pointer-events: none;
opacity: 0.9;
-webkit-box-sizing: border-box;
-moz-box-sizing: border-box;
box-sizing: border-box;
}
.bootstrap-select .no-results {
padding: 3px;
background: #f5f5f5;
margin: 0 5px;
white-space: nowrap;
}
.bootstrap-select.fit-width .dropdown-toggle .filter-option {
position: static;
display: inline;

PAGE \* MERGEFORMAT 2
 padding: 0;
}
.bootstrap-select.fit-width .dropdown-toggle .filter-option-inner,
.bootstrap-select.fit-width .dropdown-toggle .filter-option-inner-inner {
display: inline;
}
.bootstrap-select.fit-width .dropdown-toggle .caret {
position: static;
top: auto;
margin-top: -1px;
}
.bootstrap-select.show-tick .dropdown-menu .selected span.check-mark {
position: absolute;
display: inline-block;
right: 15px;
top: 5px;
}
.bootstrap-select.show-tick .dropdown-menu li a span.text {
margin-right: 34px;
}
.bootstrap-select .bs-ok-default:after {
content: '';
display: block;
width: 0.5em;
height: 1em;
border-style: solid;
border-width: 0 0.26em 0.26em 0;
-webkit-transform: rotate(45deg);
-ms-transform: rotate(45deg);
-o-transform: rotate(45deg);
transform: rotate(45deg);
}
.bootstrap-select.show-menu-arrow.open > .dropdown-toggle,
.bootstrap-select.show-menu-arrow.show > .dropdown-toggle {

PAGE \* MERGEFORMAT 2
 z-index: 1061;
}
.bootstrap-select.show-menu-arrow .dropdown-toggle .filter-option:before {
content: '';
border-left: 7px solid transparent;
border-right: 7px solid transparent;
border-bottom: 7px solid rgba(204, 204, 204, 0.2);
position: absolute;
bottom: -4px;
left: 9px;
display: none;
}
.bootstrap-select.show-menu-arrow .dropdown-toggle .filter-option:after {
content: '';
border-left: 6px solid transparent;
border-right: 6px solid transparent;
border-bottom: 6px solid white;
position: absolute;
bottom: -4px;
left: 10px;
display: none;
}
.bootstrap-select.show-menu-arrow.dropup .dropdown-toggle .filter-option:before {
bottom: auto;
top: -4px;
border-top: 7px solid rgba(204, 204, 204, 0.2);
border-bottom: 0;
}
.bootstrap-select.show-menu-arrow.dropup .dropdown-toggle .filter-option:after {
bottom: auto;
top: -4px;
border-top: 6px solid white;
border-bottom: 0;
}

PAGE \* MERGEFORMAT 2
.bootstrap-select.show-menu-arrow.pull-right .dropdown-toggle .filter-option:before {
right: 12px;
left: auto;
}
.bootstrap-select.show-menu-arrow.pull-right .dropdown-toggle .filter-option:after {
right: 13px;
left: auto;
}
.bootstrap-select.show-menu-arrow.open > .dropdown-toggle .filter-option:before,
.bootstrap-select.show-menu-arrow.show > .dropdown-toggle .filter-option:before,
.bootstrap-select.show-menu-arrow.open > .dropdown-toggle .filter-option:after,
.bootstrap-select.show-menu-arrow.show > .dropdown-toggle .filter-option:after {
display: block;
}
.bs-searchbox,
.bs-actionsbox,
.bs-donebutton {
padding: 4px 8px;
}
.bs-actionsbox {
width: 100%;
-webkit-box-sizing: border-box;
-moz-box-sizing: border-box;
box-sizing: border-box;
}
.bs-actionsbox .btn-group button {
width: 50%;
}
.bs-donebutton {
float: left;
width: 100%;
-webkit-box-sizing: border-box;
-moz-box-sizing: border-box;
box-sizing: border-box;

PAGE \* MERGEFORMAT 2
}
.bs-donebutton .btn-group button {
width: 100%;
}
.bs-searchbox + .bs-actionsbox {
padding: 0 8px 4px;
}
.bs-searchbox .form-control {
margin-bottom: 0;
width: 100%;
float: none;
}
/*# sourceMappingURL=bootstrap-select.css.map *Thank you for your business! We
look forward to working with you again.
html, body {
height: 100%;
}

.floatleft {
float: left;
}

.floatright {
float: right;
}

.alignleft {
float: left;
margin-right: 15px;
margin-bottom: 15px;
}

.alignright {
float: right;

PAGE \* MERGEFORMAT 2
 margin-left: 15px;
margin-bottom: 15px;
}

.aligncenter {
display: block;
margin: 0 auto 15px;
}

a:focus {
outline: 0px solid;
}

img {
max-width: 100%;
height: auto;
}
fix {
overflow: hidden;
}

p{
margin: 0 0 15px;
color: #444;
}

h1, h2, h3, h4, h5, h6 {

font-family: 'Raleway', sans-serif;
margin: 0 0 15px;
color: #444;
font-weight: 500;
}

PAGE \* MERGEFORMAT 2
h1 {
font-size: 48px;
line-height: 50px;
}

h2 {
font-size: 38px;
line-height: 40px;
}

h3 {
font-size: 30px;
line-height: 32px;
}

h4 {
font-size: 24px;
line-height: 26px;
}

h5 {
font-size: 20px;
line-height: 22px;
}

h6 {
font-size: 16px;
line-height: 20px;
}

a{
transition: all 0.3s ease 0s;
PAGE \* MERGEFORMAT 2
 text-decoration: none;
}

a:hover {
color: #3EC1D5;
text-decoration: none;
}

a:active, a:hover {
outline: 0 none;
}

body {
background: #fff none repeat scroll 0 0;
color: #444;
font-family: 'Open Sans', sans-serif;
font-size: 18px;
text-align: left;
overflow-x: hidden;
line-height: 22px;
}

/* Back to top button */

.back-to-top {
position: fixed;
display: none;
background: #3EC1D5;
@media (max-width: 768px) {
.back-to-top {
bottom: 15px;
}
}
PAGE \* MERGEFORMAT 2
.back-to-top:focus {
background: #3EC1D5;
color: #fff;
outline: none;
}

.back-to-top:hover {
background: #3cd6ed;
color: #fff;
}

.clear {
clear: both;
}

ul {
list-style: outside none none;
margin: 0;
padding: 0;
}

input, select, textarea, input[type="text"], input[type="date"], input[type="url"],

input[type="email"], input[type="password"], input[type="tel"], button,
button[type="submit"] {
@media (max-width: 768px) {
.back-to-top {
bottom: 15px;
}
}

.back-to-top:focus {
PAGE \* MERGEFORMAT 2
 background: #3EC1D5;
color: #fff;
outline: none;
}

.back-to-top:hover {
background: #3cd6ed;
color: #fff;
}

.clear {
clear: both;
}

ul {
list-style: outside none none;
margin: 0;
padding: 0;
}

input, select, textarea, input[type="text"], input[type="date"], input[type="url"],

input[type="email"], input[type="password"], input[type="tel"], button,
button[type="submit"] {
section-headline h2 {
display: inline-block;
font-size: 40px;
font-weight: 600;
margin-bottom: 70px;
position: relative;
text-transform: capitalize;
}

PAGE \* MERGEFORMAT 2
.section-headline h2::after {
border: 1px solid #333;
bottom: -20px;
content: "";
left: 0;
margin: 0 auto;
position: absolute;
right: 0;
width: 40%;
}

.sec-head {
display: inline-block;
font-size: 17px;
font-weight: 600;
margin-bottom: 0;
padding: 0 0 10px;
text-transform: uppercase;
transition: all 0.4s ease 0s;
}

PAGE \* MERGEFORMAT 2
 B. SCREENSHOTS

Fig. B.1: Importing Required Libraries

PAGE \* MERGEFORMAT 2
 Fig. B.2: Dataset

Fig.B.3: Dataset Information

PAGE \* MERGEFORMAT 2
Fig.B.4: Pre-processing the data

Fig.B.5: Splitting the dataset

PAGE \* MERGEFORMAT 2
 Fig.B.6: Creating the Model

Fig.B.7: Predicting the Accuracy

PAGE \* MERGEFORMAT 2
 Fig.B.8: Hash Table

C. RESEARCH PAPER

FUZZY BASED GEO-SPATIAL CRIME CATEGORY PREDICTION FOR

CRIME MAPPING

ABSTRACT: As a clean kind of crime, being evaluated. The kind of crime, time
causing issues for people and society is of day, and location all have important
seen nowadays. Increased crime leads roles. Using learning techniques like the
to increased inequality among the Random Forest Classifier, this matches
country's voters. The structure of the students into distinct classes. When
crime must be understood in order to compared to preset occupations, the
evaluate and comprehend this type of ensuing crime and caring behaviors
criminal behavior. This direction has have been shown to be relatively
given the use of open-source criminal extreme.
analysis from Kaggle in which the
prediction of the most recent crimes is KEYWORDS: K-Method, Simple Sign,
created. This project's key factor is Prediction Fee Regression, association,

PAGE \* MERGEFORMAT 2
rule mining etc. attempted to be repeated in the same
case. In criminal cases, it depends on
INTRODUCTION many factors like the intelligence of the
criminals, security of the area, work
Every day, the crime rate is being done on the clues used in the data.
observed as cutting-edge technology Data collection, data classification,
and high-tech techniques aid criminals pattern recognition, prediction, and
in reaching their goals. However, visualization are encompassed by
according to the Crime Bureau, crimes critical steps in analysis. The proposed
such as theft, arson, murder, rape, and framework utilizes various visualization
cruelty have increased, as has group techniques to depict crime trends and
betting. Crime- related information is employs different methods for
gathered from multiple blogs, predicting crime. Machine learning
information, and websites. Creating a algorithms are utilized, and time (hour,
database of crime reports necessitates day, month, and year), location (latitude
the utilization of massive volumes of and longitude), and type of crime
data as records. Crime rates can be constitute the input data for our
lessened by knowledge received algorithm are :
through machine learning techniques. -Robbery
Quick detection of criminals and crime- -Rape
prone locations is possible. Faster -Theft
enablement of pay payments and the -The Killing
provision of accurate consequences are -Abduction
facilitated by machine learning when It is possible that some criminal act may
applied to crime datasets, which are have been committed due to this.
informative with machine learning Various types of algorithms, including
techniques. A unique way to discover random forests, are tested. Several
crime hotspots has been found by law additional tasks are performed: first, an
enforcement officials. Regions where attempt is made to guess which of the 6
crime is fairly focused are alluded to. types of crimes occurred most often,
Better results in the fight against crime and then an attempt is made, which
reporting datasets can be achieved includes violence and violent crimes.
through the utilization of machine
learning techniques. The first step in the OBJECTIVE
study of crime is represented by the
analysis. Through the exploration,  Our paintings are ceased.
connection, and revelation of  Before the crime is taken place.
relationships in crime research,  Crime scenes are provided.
statistics, queries, and reports on  Crime is modelled.
demand are generated. This is how it  Crimes are described consistent
helps to determine if payment is needed with the place where they arise.
for a particular known item or a new
model. Crimes, as recommended RELATED WORK
actions by active criminals, may be
referred to as such. If so, the offense is A literature evaluation is considered to
PAGE \* MERGEFORMAT 2
be an essential step in the software crime from statistics to better
development process. Portrayed are comprehend criminal behaviour, and,
two test cases given, and each case is combined with records analysis, the
considered, outlining the outcomes that prevention of future crimes is targeted.
empower elements of the organization. Although many mining techniques use
The subsequent stage is concluded, rule mining and affiliation strategies,
determining which working system and fewer tactics are focused on crime
apparatus language can be utilized to prediction models. In this article,
grow the gadget once those conditions various sorts of LSOA (Lower Layer
are met. Massive external support is over Output Area) models predicting the
required when an instrument is being frequency of crime and delinquent
created by software engineers, and this behaviour are examined. A variety of
assistance can be obtained from senior methods, such as learning, regression,
packages, books, or websites. When and decision trees, are employed with
planning the objectives of the system three algorithms. The records,
before building the machine, the above consisting of over 600,000 records
points are kept in mind. The before processing, are provided by UK
development scope is considered by police forces. The results indicate that
most businesses, and a thorough the recurrence of a specific type of
evaluation of everything necessary for crime and delinquent behaviour can be
the project's development is conducted. anticipated using choice trees (M5P
Documentation review is considered the algorithm), considering prediction
most essential part of the software performance and processing time.
development process for any reason.
The factors, resource requirements, 2. Wrongdoing Examination and
manpower, finance, and strengths of Forecast Utilizing Information Mining
the company are recognized and
analyzed before the equipment is Crime evaluation and prevention are
developed and incorporated into the techniques for identifying and analyzing
project. After thoroughly and carefully criminal patterns and developments.
checking all of these parameters, the Regions with an excessive possibility of
specification of the software on the crime can be predicted, and crime-
concerned computer is decided, susceptible regions can be visualized
according to which type of operating by our gadget. The method involved in
system is needed for the purpose, and tackling wrongdoings can be
all the necessary software is required. accelerated by cops through crime data
The development of the system and evaluation using computerized systems.
associated functions, along with a Previously unknown useful information
degree, is carried forward. can be extracted from unstructured
information by the possibility of insights
1. Predictive Crime Research via mining. It is a technique between pc
Open Data Mining technology and criminal justice that
makes information mining assist in
It involves an increase in study methods solving wrongdoings more quickly.
aimed at extracting an understanding of Often, the day-to-day elements of crime
PAGE \* MERGEFORMAT 2
are focused on rather than the reasons
for wrongdoing, alongside criminal Strategies. The process of generating
climate and political hostility. Crime critical data for a company through data
Detection Techniques use Data Mining mining involves evaluating and
and K-Means. analyzing large pre-existing databases.
Extracting new information from existing
3. Solving Procedures Utilizing datasets is part of the predictions made.
Information Mining and K -Media The application of various analytical
and inferential strategies in criminology
If crime is repeated within the society, involves using information mining, with
institutions and businesses will be significant work conducted. Integration
impacted in some way or another. of all this information has been
Therefore, it seems essential to attempted. Police departments and
examine the reasons, factors, and similar criminal justice organizations
interrelationships of various crime maintain numerous large databases for
phenomena with the purpose of predicting or analyzing criminal activity
discovering the most appropriate and behaviour within the community.
methods to control and prevent crime. Utilizing crime information to identify
The reasons, factors, and criminals is also done. Efforts have
interrelationships of various been made to integrate this information.
The examination of crime phenomena The primary objective of this work is to
aims to uncover the most suitable investigate the learning techniques,
methods for controlling and preventing both visible and invisible, employed in
crime. The essential objective of this discovering criminals. This paper
paper is to have association violations presents a study of criminology and
separated according to the recurrence prediction using multiple mining
in their commonness in particular years. techniques.
Information mining is comprehensively
used to explore, inspect, and select 5. Criminal Example Examination,
examples of various wrongdoings. A Representation and Expectation
hypothetical model was executed based Utilizing Information Mining
on statistics mining methods, including
classification and type, on genuine Concerns about crimes against women
presence wrongdoing data kept by are present in all countries worldwide
police in Britain and Grains between today. Many countries are making
1990 and 2011. The model's quality efforts to address this issue.
was enhanced and low values were Implementation of preventive measures
eliminated by weighting the abilities. is underway to decrease crimes against
The parameters of outlier detection women. A significant amount of
were optimized using a genetic information is generated from crime
algorithm (GA) and its Quick Miner tool. reports annually. This information is
highly useful for studying and predicting
4. Research the usage of facts on crimes, contributing to the prevention of
crime evaluation and information crimes to some extent. The analysis of
mining crime is a crucial part of police
PAGE \* MERGEFORMAT 2
department work. Crime information understated. The indirect connection
analysis can help examine crime between the crimes is significant. This
patterns, clues to interrelationships, etc. is why statistics mining can be very
The significance of the indirect beneficial in solving, visualizing, and
connection between crimes cannot be predicting crimes using crime datasets.
Grouping and correlating records make information mining. Rather than
it easier to identify similarities and focusing on the causes of crime, such
differences between data objects. The as criminal climate and political hostility,
group data item is used as a table. The the emphasis is primarily on the daily
dataset is specified in a predefined factors of crime.
state. A link has been provided,
encompassing various types of crimes 7. A Systematic evaluate of crime
against women in different states and statistics collection
urban communities of India. Crime
planning control can assist in planning Criminal patterns and tendencies are
crime prevention strategies, and, Data systematically identified and analyzed
mining methods assist in predicting and through criminology. With the
visualizing information in various advancement of computer systems, the
formats for a better understanding of process of solving crimes can be
crime patterns. Crime analysis utilizes expedited by law enforcement with the
data mining techniques. This kind assistance of crime data analysts.
information Records can be examined for
previously unknown valuable
6. Wrongdoing Examinationand information through the utilization of the
Expectation Utilizing Information concept of data mining. The
Mining Techniques identification of criminals is facilitated by
predictive policing, employing analytical
The method employed in crime techniques and predictive strategies,
assessment and prevention involves proving highly effective. Due to the
identifying and analyzing criminal increased crime rates in recent years,
patterns and trends. Anticipation of the challenge of handling large
areas with a high probability of crime quantities of crime statistics stored in
can be facilitated, and regions prone to warehouses arises, which can be very
wrongdoing can be visualized by our challenging to manually analyze. The
system. With the advent of automated police are urged to take a proactive
systems, the process of solving crimes stance. An evaluation of the methods
can be expedited by crime data analysis and techniques used for the
using computerized methods. The identification of criminals is presented in
application of the concept of data this article.
mining allows previously unknown
valuable information to be extracted 8. A Thesis on Crime Prediction via
from unstructured data. It is a technique Holistic Approach
between computer science and criminal
justice that enables crimes to be solved The paper explores several data mining
more quickly with the assistance of and learning algorithms utilized in data

PAGE \* MERGEFORMAT 2
mining. An overview of crime analysis because the classifier uses explicit
and forecasting techniques is presented values, which obscures the results.
in this dissertation. The technique of For mentioned names with higher cost
reducing future crimes by identifying results. The methods are not suitable
and predicting them is known as crime for classifying regions with real data
forecasting. Historical data is used in and features. The values of the
the practice of crime forecasting, and, classifier must be in range, so it is
after analyzing the information, future essential to specify the best value.
crime is predicted based on region, System architecture design is the plan
time, day, season, and year. The way of characterizing the subsystems
number of crimes is increasing rapidly, that make up the device and the control
making it essential to more accurately and conversation structure of the
predict future crimes. Solving the subsystems. The reason of an
problem of crime by discovering hidden architect's mission is to create the
crime patterns is greatly facilitated by overall structure of a software device.
data mining technology. Therefore, the
aim of this study is to have the various PROPOSED SYSTEM
analyzed and discussed. techniques
used for crime forecasting and analysis.
A rational analysis of data mining and
predictive type techniques for detecting
future crimes is presented in this paper.
Filters and machine learning layers are
EXISTING SYSTEM first employed to preprocess the
obtained data, removing redundant and
In pre-processing, the dataset acquired duplicate values, while also reducing
from an open source is initially pre- dimensionality for cleaner data.
processed to eliminate existing Subsequently, a fragmentation process
duplicate values. A decision tree is is applied to the extra data, dividing it
used to extract features from large into testing and training datasets. The
amounts of data and identify crime dataset is utilized for both training and
patterns. It provides a primary testing the model. The objective is to
framework for the process of further transform the type of crime, year,
classifying the included data. Crime month, time, date, and region into
patterns are identified using deep integers for easy classification. A
neural networks. Performance, as random jump classifier is initially used
expected, computer-trained and tested to generate independent influence
quotes. Crime forecasting helps predict between the features. This
any type of future, solving crimes as characterization of crimes enables the
quickly as possible, and supporting the analysis of crime occurrences at
government. specific times and places. Ultimately,
crimes that are spatially and temporally
Disadvantages of Existing System:
frequent are identified. The model's
information is filled by calculating
Previous works have low accuracy
accuracy. Python is the language
PAGE \* MERGEFORMAT 2
utilized for developing predictive from Kaggle and other resources.
models, and it executes analytical and
machine-learning models. Records Series :

Advantages of Proposed System : Let the data be exchanged. Missing

records will be removed, and a list of
When dependent on multiple features,
segment names that are desired to be
the identification of crime patterns is
saved or kept will be created. Then, all
found to be more suitable by the
segments other than those that are
proposed algorithm. Date and place.
wanted to be saved will be erased or
This solves the problem of analyzing
removed. Finally, lines with missing
the impact of independent features. The
qualities will be eliminated from the
introduction of the optimal value is not
dataset.
required as it considers the actual
nominal value and factors with
Pattern Analyzing :
adequate knowledge of the area. The
accuracy was high compared to other
When machine learning training is
models of predictive learning tools.
employed, two sets of data are needed:
BLOCK DIAGRAM one for demonstration and one for
testing. However, now only one set is
The system architecture design is the available. Therefore, let it divided into
planned way of defining the subsystems parts in the ratio of 80:20. Let the data
that make up the system and the frame be split into feature columns and
control and communication structure of label columns. It was passed through
the subsystems. The purpose of an training from Sklaren. Utilize that data
archite. for the establishment of the partition.
Set the test size to 0.2, with 80%
Fig - 1 allocated to the training set and 20% to
the testing set. After the model is
MODULES AND THEIR created, testing is deemed crucial. To
DESCRIPTION validate this, the test_x method is
employed.
Information Series
PROPOSED ALGORITHM
The first genuine step towards
constructing a machine learning model Random Forest: A classifier algorithm
for data series is taken. This is a crucial that utilizes many Tree Selections in
step that depends on an excellent Distributed Data Collection and
model, and the bigger and better it is, Application Generic for a wider
the better our adaptation will perform. prediction records set precisely is
Various procedures of data collection, employed. A random forest is created in
such as text scraping and manual two steps: Firstly, an uneven wooded
intervention, are employed. Machine area is formed with a useful blend of
learning algorithms are compared for resources, and N trees are defined,
crime scene prediction, utilizing data after which it declares itself as a tree
PAGE \* MERGEFORMAT 2
created in the first degree. The works of is received by a majority vote. In our
the developers are thus explained. research, when implementing the K-NN
Steps and map: K knowledge elements algorithm in current systems, it is found
to be less accurate compared to other
systems implementing various
mechanisms.

Decision Tree: A tree of judgment is a

type of tree, shaped like a flowchart,
where each internal node represents a
personal test result, and leaf nodes
Album exercise are randomly picked. represent class topics or performance.
Checkboxes are created in access Dozens more, the solution can process
standards, and subsets are chosen. hundreds of measurements trees. The
Another N is chosen for the bushes to feature is checked to insert the search
be created. Beyond 1 & Again II. given the root node of the tree. Related
Expenditure declared by the desire tree to the type of tree, the sub-tree to
for new ones. Information is provided to assign the cost is placed at the new
the new variable Helper with the node, and then the same could be
maximum votes. This algorithm is resubmitted. Crime data is examined
usually better than how the outcome is and predicted using attributes like
predicted. Various studies have shown longitude (X), latitude (Y), address,
that predicting crimes at random is Saturday, day (YYYY-mm-
more accurate. Wood is used. The dd:hh:MM:ss), area, clearance, and
biggest disadvantage is that a random type. It is said correctly, and to avoid
forest can be slow and may be delayed overcrowding, exploration of numerous
when there are many trees. Trained programs is moved on to the main
algorithms expect very slowly but evaluation.
practice quickly. Fuzzy safety route project accuracy
KNN train: 98%. Test: 94%
KNN: Essentially, an attempt is made to DT train: 99%. Test: 98%
locate similarities in erasing information RF train: 99% Test: 98%
between testing and setup. If the rail is
installed and the test kit is close, the
testing kit is compared to the training
dataset. This is the simplest downside Graph:
of the algorithm. It becomes challenging
when setting up a small number in EXPERIMENTAL ANALYSIS
training data logs. To remedy this Confusion Matrix
hassle, the K-NN algorithm is used. The The performance of a set of test data is
K- NN algorithm is assumed by the new summarized by a confusion matrix
case and the present case to be employed in machine learning models.
comparable, and a new instance of a Within this matrix, instances from the
very similar kind is presupposed to model's predictions, both accurate and
exist, more precisely, that the outcome inaccurate, are displayed—a means of
PAGE \* MERGEFORMAT 2
evaluating the model's performance.
Commonly used in assessing
classification models aiming to predict a
categorical label for each input
instance, the instances produced by the For the above case:
model on the test data are revealed. Recall = 8 / (8+1) =8/9 = 0.8888
True positives (TP) manifest when a F1-Score
positive data point is accurately F1 score in machine learning measures
predicted by the model, and true precision and recall balance as
negatives (TN) emerge when a harmonic mean. Ranges from 0 to 1
negative data point is accurately and evaluates binary classifiers
predicted. False positives (FP) occur effectively
when a positive data point is incorrectly
predicted by the model, while false
negatives (FN) arise when a negative
For the above case:
data point is incorrectly predicted.
F1-Score: = (2* 0.8888*
Accuracy, a measure of the model's
0.8888)/( 0.8888+ 0.8888) = 0.8888
performance, is determined by the ratio
Table:
of total correct instances to total
precision recall f1- support
instances. score
0 1.00 1.00 1.00 86
1 0.82 0.74 0.78 19
2 0.97 0.98 0.97 147
3 1.00 1.00 1.00 140
For the above case: 4 1.00 1.00 1.00 19
Accuracy = (8+6)/ (8+6+1+1) = 14/16 = 5 1.00 1.00 1.00 3
0.875
Precision Training:
The definition of accuracy with a The predictive models, including KNN,
formula has been observed, and now Decision Tree (DT), and Random
the definition of precision using a Forest (RF), were assessed during the
formula is going to be examined. training phase. Optimal performance for
KNN and DT was revealed as error
rates were compared across different K
values.
For the above case:
Precision = 8/ (8+1) =8/9 = 0.8888
Recall
Recall assesses a classification model's
efficacy in recognizing all pertinent
instances. Calculated by TP/(TP+FN), it
gauges the model's capability to identify
positives. High recall: effective; low
recall: misses positives. Fig-3
Testing:
PAGE \* MERGEFORMAT 2
In testing Critical insights were revealed
through the analysis of feature
importance. The impact of various
factors on crime prediction was
assessed by the model, emphasizing
key features and their relative
importance in determining accurate
outcomes for crime mapping. Fig-7

Fig-8

Fig-4

RESULT AND DISCUSSION

Fig-9

Fig-5

Fig-10

With the assistance of machine learning

technology, relationships and patterns
Fig-6
between unique information have
become easier to identify. The primary
aim of predicting the type of crime that
will occur when the location of the crime
is known is served by this software.
Using machine learning concepts, a
model was developed by us utilizing the
PAGE \* MERGEFORMAT 2
dataset, cleaning, and transforming the analysis that can help discover it more
information. efficiently and effectively.

CONCLUSION REFERENCES

Crimes occur in various categories, 1.Ginger Saltos and Mihaela Coacea,

ranging from small cities to large cities - An Exploration of Crime prediction
Robbery, Murder, Kidnapping, Assault, Using Data Mining on Open Data,
Battery, False Imprisonment, International journal of Information
Kidnapping, and Manslaughter. As soon technology & Decision Making,2017.
as a crime occurs, cases need to be 2.Shiju Sathyadevan, Devan M.S,
promptly resolved. In Input it is reported Surya Gangadharan.S, Crime Analysis
that crimes precise location details are and Prediction Using Data Mining, First
encompassed by geospatial data, International Conference on networks &
utilizing latitude and longitude soft computing (IEEE) 2014.
information, with temporal data 3.Khushabu A.Bokde, Tisksha
factoring in the date and time. In P.Kakade, Dnyaneshwari S. Tumasare,
Output, Forecasted crime categories Chetan G.Wadhai B.E Student, Crime
are generated by the system, predicting Detection Techniques Using Data
probable crime classifications in a given Mining and K- Means, International
geospatial context. Journal of Engineering Research &
technology (IJERT) ,2018.
FUTURE ENHANCEMENT 4.H.Benjamin Fredrick David and
A.Suruliandi,Survey on crime analysis
In the future, crime record units can be and prediction using data mining
subjected to assessment using the techniques, ICTACT Journal on Soft
same models, which will employ more computing, 2017.
advanced algorithms to identify trends 5.Tushar Sonawanev, Shirin Shaikh,
in their predictive performance and rahul Shinde, Asif Sayyad, Crime
improve object reputation. It is believed Pattern Analysis, Visualization And
that greater accuracy can be achieved prediction Using Data Mining, Indian
by incorporating more features in the Journal of Computer Science and
address field to create a comprehensive Engineering (IJCSE), 2015.
predictive framework, enabling the 6.RajKumar.S, Sakkarai Pandi.M,
quick management of resources by law Crime Analysis and prediction using
enforcement agencies in a specific data mining techniques, International
area. To determine the quantity and Journal of recent trends in engineering
intensity of crime activity, a more & research,2019.
thorough temporal analysis can be 7. Sarpreet kaur, Dr. Williamjeet Singh,
conducted through time series analysis, Systematic review of crime data mining
a combination of temporal and spatial

PAGE \* MERGEFORMAT 2
International Journal of Advanced
Research in computer science, 2015.#
DEPARTMENT OF COMPUTER
SCIENCE AND ENGINEERING
Ayisheshim Almaw, Kalyani Kadam,
Survey Paper on Crime Prediction
using Ensemble Approach, International
journal of Pure and Applied
Mathematics,2018.

PAGE \* MERGEFORMAT 2
8.Dr .M.Sreedevi, A.Harha Vardhan
Reddy, ch.Venkata Sai Krishna Reddy,
Review on crime Analysis and
prediction Using Data Mining
Techniques, International Journal of
Innovative Research in Science
Engineering and technology ,2018.
9. K.S.N .Murthy, A.V.S.Pavan kumar,
Gangu Dharmaraju, international
journal of engineering, Science and
mathematics, 2017.
10. Deepiika k.K, Smitha Vinod, Crime
analysis in india using data minig

PAGE \* MERGEFORMAT 2