MAHENDRA INSTITUTE OF TECHNOLOGY

(Autonomous)
Mahendhirapuri,Mallasamudram,Namakkal DT-637503

DEPARTMENT OF INFORMATION TECHNOLOGY

CRIME TYPE AND OCCURRENCE PREDICTION USING MACHINE LEARNING

A FINAL YEAR PROJECT REVIEW
Presented by

ARUL SELVAN V 611620501022

KALAIVANAN K 611620501022
RAJESHKUMAR M 611620501022
SANTHOSH R 611620501022

Date : 06.06.2024
Guided by,
Mr. K.DHIVAKAR M.E.,
Assistant Professor
Department of Information Technology
 Abstract

• The ability to analyze this amount of data with its inherent complexities without using computational support puts a strain on
human resources.

• This examines the current techniques that are used to predict crime and criminality. Over time, these techniques have been
refined and have achieved limited success.

• Machine Learning is the procedure which includes evaluating and examining large pre-existing databases in order to generate
new information which may be essential to the organization.

• The criminals can also be predicted based on the crime data.

• The main aim of this work is to perform a survey on the supervised learning and unsupervised learning techniques that has
been applied towards criminal identification.

• This project presents the survey on the Crime analysis and crime prediction

2
 Existing System
• Integrates and reduces the extracted crime data into structured 5,038 crime instances.

• We represent these instances using 35 predefined crime attributes. Safeguard measures are taken for the crime
database accessibility.

• Rest four modules are useful for crime detection, criminal identification and prediction, and crime verification,
respectively.

• Payload attribution systems (PAS) are one of the most important tools of network forensics for detecting an offender after
the occurrence of a cybercrime.

• A PAS stores the network traffic history in order to detect the source and
destination pair of a certain data stream in case a malicious activity occurs on the network .

• Our approach contributes in the betterment of the society by helping the investigating agencies in crime detection and
criminals’ identification, and thus reducing the crime rates.
3
 DRAWBACK OF EXISTING SYSTEM

1. Reduction in Accuracy

2. Miss Classification

3. Less Prediction

4
 Proposed System
• In proposed a tool which applies an enhanced Random forest Algorithm to detect the suspicious e-mails about the
criminal activities.

• An improve with an enhanced feature selection method and attribute-importance factor is applied to produce a better
and faster Random forest algorithm based on the information entropy which is explicitly derived from a series of
training data sets from several classes.

5
 Advantages
• To Conducting criminal analysis and identifying trends in crime.

• Disseminate knowledge to help with the creation of crime reduction and preventive measures.

• Recognize and examine recurring criminal trends to prevent similar incidents from happening again.

6
 HARDWARE REQUIREMENTS

• Ram - 4 GB

• Hard Disk - 80 GB

• Processor- Above i3 10th Gen

7
 SOFTWARE REQUIREMENTS

• OS - Windows 10 OS 64 bit

• Language - PYTHON

• Tool - Anaconda navigator

8
 MODULES

• Training Data

• Testing Data

• Pre –Processing

• Feature Extraction

• Normalization

• Random Forest Algorithm

9
 TRAINING DATA

• The quality, variety, and quantity of your training data determine the success of your machine learning
models.

• The form and content of the training data often referred to as labeled or human labeled data or ground truth
dataset is designed for to train specific ML models with an end application in perspective

10
 TESTING DATA

• Test data is data which has been specifically identified for use in tests, typically of a computer program.

• Some data may be used in a confirmatory way, typically to verify that a given set of input to a given
function produces some expected result.

11
 PRE –PROCESSING

• Data preprocessing which mainly include data cleaning, integration, transformation and reduction, and
obtains training sample data needed.

• It is a data mining technique that transforms raw data into an understandable format

12
 FEATURE EXTRACTION

• Feature extraction is the process of transforming the input data into a set of features which can very well
represent the input data.

• It is a special form of dimensionality reduction

13
 NORMALIZATION

• Normalization is the process of scaling individual samples to have unit norm.

• This process can be useful to use a quadratic form such as the dot-product or any other kernel to quantify
the similarity of any pair of samples

14
 RANDOM FOREST ALGORITHM

• It is the basic classifier and it establishes a large number of trees.

• Random forests is an effective prediction tool widely used in data mining.

• It constructs a series of classification trees which will be used to classify a new example.

• The idea used to create a classifier model is constructing multiple decision trees, each of which uses a subset
of attributes randomly selected from the whole original set of attributes.

15
 Conclusion

• This is a theoretical study for several methods in identification of crime by the random forest algorithm. give better
accuracy.

• we have examined the accuracy of class and prediction based totally on different check sets.

• Classification is done based on the Bayes theorem which showed more than 90% accuracy.

• Using this algorithm we trained numerous news articles and build a model.

• For testing we are inputting some test data into the model which shows better results.

16
 References

• A Forensic Investigation Tool for Identifying the Influential Members

of a Criminal Organization”, IEEE Transactions on

• Evidence-based Analysis of Mentally 111 Individuals in the Criminal Justice System”, Proceedings of IEEE
Systems and Information Engine

• “ Mining Location-based Social Networks for Criminal Activity Prediction”, Proceedings of 24th IEEE
International Conference on Wireless and Optical Communication,

17
 THANK YOU 

18