CYBER THREAT
DETECTION PLATFORM
USING MACHINE
LEARNING
TEAM MEMBERS
ARUNKUMAR.T- 513320104002
GANESH KUMAR.P- 513320104011
MUTHU KRISHNAN.S-513320104026
GUIDE
DR.J.SANTHANA KRISHNAN
CSE DEPARTMENT
UCEA
INTRODUCTION
• In todays’s interconnected world,cyber threats pose a significant risk to
organization of all sizes.
• Increase reliance on digital technologies for conducting business
operations,storing sensitive data and communicating with customers,the potential
impact of cyber threats has grown significantly.
• From data breaches and financial losses to reputational damage and regulatory
fines,organizations face a wide range of consequences.
• Crucial for safegaurding against these risks and ensuring the resilience
of businesses in today’s interconnected world.
ABSTRACT
• Cyber Threat Detection platform using machine learning is to detect and address
the malicious URL’s.
• The objective of Malicious website is to fraud or steal the personal or financial
details of users.
• We classify the Raw URL’s into different class types such as benign
URL’s,Phishing URL’s,malware URl’s or defacement URL’s
• Benign-Kind of Safe URL’s.
• Phishing-Sending fradulant communications and spam messages .
• Malicious Defacement-Unauthorized alteration of existing content on website.
• Malware URL’s-injecting a malware into a victim’s system.
EXISTING SYSTEM
• Malicious Web sites are the basis of most of the criminal activities over the internet.
• The dangers that arise due to the malicious sites are enormous and the end-users must be
prohibited from visiting such sites.
• The users should prohibit themselves from clicking on such Uniform Resource Locator (URL).
• In order to prevent such attacks, the paper proposes the use of machine learning algorithms to
detect
• Phishing Websites. The Existing PWD (Phishing Website Detection) model is trained using an
existing dataset which contains URLs, each with unique features, and is applied to three different
• machine learning classififiers support vector machine, logistic regression and Naive Bayes. After
training and testing the algorithms, it is observed that Naive Bayes classififier recorded the
highest accuracy
DISADVANTAGES
• Low Accuracy Due to Training Loss
• Many Website features not included for the consideration
S.NO Title Publication and Year Advantages Disadvantages
1 Phishing or not phishing?
A Survey on the
Detection of Phishing
Websites.
2023 University of
Pavia Through the
CRUI-CARE
Agreement
depicts higher accuracy
and f1 score in phishing
detection
is less prone to overfitting
compared to individual
decision tree
2 The Emergence Threat of
Phishing attack and the
detection technique using
machine learning models
2021 International
Conference on
automation,control and
mechatronics for
industry.
Accurately classifying
both phished and non
phished url’s.
Model is not properly tuned
and the dataset is too small.
3 Phishing Attack Detection
on Text Messages using
Machine Learning
Techniques
2022 IEEE Pune
Section International
Conference(PuneCon)
phished messages
getting correctly
predicted and the
machine earning
teachnique to build
model is handles noisy
data well
it may lack interpretability
compared to simpler models
LITERATURE SURVEY
PROPOSED SYSTEM
• Collect dataset containing benign,phishing and malicious websites from the open
source platforms.
• Divide the dataset into training and testing sets.
• Analyze and preprocess the dataset by using EDA techniques.
• Compare the obtained results for trained models and specify which is better.
• We need to Detect and identify the different types of threats such as benign
URL’s,Phishing URL’s,malware URL’s and defacement URL’s
• To increase the accuracy of model detection Random Forest Classifier,Light GBM
and XGBoost are need to implement the machine learning technique.
REQUIREMENTS
• Python3
library files: pandas,itertools,numpy,seaborn
vurtual environment: Anaconda
• Macine learning Technique
Algorithm:Random Forest Classifier,XGBoost,Light GBM
REFERENCE
• 1.Phishing Attack Detection on Text Messages using Machine Learning
Techniques - 2022 IEEE Pune Section International Conference(PuneCon)
• 2.The Emergence Threat of Phishing attack and the detection technique using
machine learning models -2021 International Conference on automation,control
and mechatronics for industry.
• 3.Phishing or not phishing? A Survey on the Detection of Phishing Websites-2023
University of Pavia Through the CRUI-CARE Agreement.
Thank You

CYBER THREAT DETECTION PLATFORM USING MACHINE LEARNING.pptx

  • 1.
    CYBER THREAT DETECTION PLATFORM USINGMACHINE LEARNING TEAM MEMBERS ARUNKUMAR.T- 513320104002 GANESH KUMAR.P- 513320104011 MUTHU KRISHNAN.S-513320104026 GUIDE DR.J.SANTHANA KRISHNAN CSE DEPARTMENT UCEA
  • 2.
    INTRODUCTION • In todays’sinterconnected world,cyber threats pose a significant risk to organization of all sizes. • Increase reliance on digital technologies for conducting business operations,storing sensitive data and communicating with customers,the potential impact of cyber threats has grown significantly. • From data breaches and financial losses to reputational damage and regulatory fines,organizations face a wide range of consequences. • Crucial for safegaurding against these risks and ensuring the resilience of businesses in today’s interconnected world.
  • 3.
    ABSTRACT • Cyber ThreatDetection platform using machine learning is to detect and address the malicious URL’s. • The objective of Malicious website is to fraud or steal the personal or financial details of users. • We classify the Raw URL’s into different class types such as benign URL’s,Phishing URL’s,malware URl’s or defacement URL’s • Benign-Kind of Safe URL’s. • Phishing-Sending fradulant communications and spam messages . • Malicious Defacement-Unauthorized alteration of existing content on website. • Malware URL’s-injecting a malware into a victim’s system.
  • 4.
    EXISTING SYSTEM • MaliciousWeb sites are the basis of most of the criminal activities over the internet. • The dangers that arise due to the malicious sites are enormous and the end-users must be prohibited from visiting such sites. • The users should prohibit themselves from clicking on such Uniform Resource Locator (URL). • In order to prevent such attacks, the paper proposes the use of machine learning algorithms to detect • Phishing Websites. The Existing PWD (Phishing Website Detection) model is trained using an existing dataset which contains URLs, each with unique features, and is applied to three different • machine learning classififiers support vector machine, logistic regression and Naive Bayes. After training and testing the algorithms, it is observed that Naive Bayes classififier recorded the highest accuracy
  • 5.
    DISADVANTAGES • Low AccuracyDue to Training Loss • Many Website features not included for the consideration
  • 6.
    S.NO Title Publicationand Year Advantages Disadvantages 1 Phishing or not phishing? A Survey on the Detection of Phishing Websites. 2023 University of Pavia Through the CRUI-CARE Agreement depicts higher accuracy and f1 score in phishing detection is less prone to overfitting compared to individual decision tree 2 The Emergence Threat of Phishing attack and the detection technique using machine learning models 2021 International Conference on automation,control and mechatronics for industry. Accurately classifying both phished and non phished url’s. Model is not properly tuned and the dataset is too small. 3 Phishing Attack Detection on Text Messages using Machine Learning Techniques 2022 IEEE Pune Section International Conference(PuneCon) phished messages getting correctly predicted and the machine earning teachnique to build model is handles noisy data well it may lack interpretability compared to simpler models LITERATURE SURVEY
  • 7.
    PROPOSED SYSTEM • Collectdataset containing benign,phishing and malicious websites from the open source platforms. • Divide the dataset into training and testing sets. • Analyze and preprocess the dataset by using EDA techniques. • Compare the obtained results for trained models and specify which is better. • We need to Detect and identify the different types of threats such as benign URL’s,Phishing URL’s,malware URL’s and defacement URL’s • To increase the accuracy of model detection Random Forest Classifier,Light GBM and XGBoost are need to implement the machine learning technique.
  • 8.
    REQUIREMENTS • Python3 library files:pandas,itertools,numpy,seaborn vurtual environment: Anaconda • Macine learning Technique Algorithm:Random Forest Classifier,XGBoost,Light GBM
  • 9.
    REFERENCE • 1.Phishing AttackDetection on Text Messages using Machine Learning Techniques - 2022 IEEE Pune Section International Conference(PuneCon) • 2.The Emergence Threat of Phishing attack and the detection technique using machine learning models -2021 International Conference on automation,control and mechatronics for industry. • 3.Phishing or not phishing? A Survey on the Detection of Phishing Websites-2023 University of Pavia Through the CRUI-CARE Agreement.
  • 10.