Nothing Special   »   [go: up one dir, main page]
Scribd Logo
Upload

Welcome to Scribd!

  • 269 views

    Whale Optimization Algorithm

    Uploaded by

    abc def
    The document summarizes a seminar presentation on credit card fraud detection using a whale optimization algorithm (WOA) optimized backpropagation (BP) neural network. It introduces the topic, outlines the contents which include literature review, implementation details of the WOA and BP neural network, applications, advantages and disadvantages. It then presents the results showing the WOA-BP model has higher fraud detection rates than the BP model alone. It concludes the WOA-BP network has stronger generalization capabilities and improved accuracy for fraud detection.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd

    Whale Optimization Algorithm

    Uploaded by

    abc def
    269 views16 pages
    The document summarizes a seminar presentation on credit card fraud detection using a whale optimization algorithm (WOA) optimized backpropagation (BP) neural network. It introduces the topic, outlines the contents which include literature review, implementation details of the WOA and BP neural network, applications, advantages and disadvantages. It then presents the results showing the WOA-BP model has higher fraud detection rates than the BP model alone. It concludes the WOA-BP network has stronger generalization capabilities and improved accuracy for fraud detection.

    Original Description:

    A machine learning algorithm based on the prey hunting nature of whale

    Original Title

    Whale optimization algorithm

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document summarizes a seminar presentation on credit card fraud detection using a whale optimization algorithm (WOA) optimized backpropagation (BP) neural network. It introduces the topic, outlines the contents which include literature review, implementation details of the WOA and BP neural network, applications, advantages and disadvantages. It then presents the results showing the WOA-BP model has higher fraud detection rates than the BP model alone. It concludes the WOA-BP network has stronger generalization capabilities and improved accuracy for fraud detection.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    269 views16 pages

    Whale Optimization Algorithm

    Uploaded by

    abc def
    The document summarizes a seminar presentation on credit card fraud detection using a whale optimization algorithm (WOA) optimized backpropagation (BP) neural network. It introduces the topic, outlines the contents which include literature review, implementation details of the WOA and BP neural network, applications, advantages and disadvantages. It then presents the results showing the WOA-BP model has higher fraud detection rates than the BP model alone. It concludes the WOA-BP network has stronger generalization capabilities and improved accuracy for fraud detection.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    of 16

    ANGADI INSTITUTE OF TECHNOLOGY &

    MANAGEMENT
    Department of Computer Science Engineering
    Seminar on
    “CREDIT CARD FRAUD DETECTION BASED ON WOA BP
    NEURAL NETWORK ”

    Project Guide:
    Prof. Bharati Kale.

    Presented by:
    Priyanka Kiran Patil 2AG15CS055
    CONTENTS
    1. Introduction

    2. Problem Statement & Motivation

    3. Literature Survey

    4. Implementation

    5. Applications

    6. Advantages & Disadvantages

    7. Result

    8. Conclusion
    INTRODUCTION
     The use of credit card is becoming more and more widespread.

     Currently decision tree algorithm, neural network algorithm are being used.

     BP neural network have drawback such as slow convergence speeds and network are
    vulnerable to local minima.

     Whale algorithm is used to optimize the initial network weights and threshold to
    improve the accuracy.
    PROBLEM STATEMENT
    A credit card fraud detection technology based on whale algorithm optimized BP
    neural network aiming at solving the problems of slow convergence rate, easy to
    fall into local optimum, network defects and poor system stability derived from BP
    neural network. Using whale swarm optimization algorithm to optimize the weight
    of BP network.

    MOTIVATION
    Detecting credit card fraud has become a task in the technically growing world.
    With the growth in technology cashless payments are also increasing. Hence to
    handle all this online as well as offline credit card transactions we use several
    methods like decision tree, neural network etc. to increase the speed and efficiency
    of detecting fraud here we can use whale optimization algorithm to assign weights
    to the neural network.
    LITERATURE SURVEY

    Sr No. TITLE AUTHOR SUMMERY


    YEAR
    Fraud detection using neural Raghvendra Patidar et.al Supervised learning feed
    1 network Year: 2011 forward back propogation
    algorithm.

    Fraud detection based on artificial Neda Soltani et.al  User behaviour is


    2 immune system Year: 2012 considered
     Two methodology,
    tracking account
    behaviour & general
    thresholding.
    LITERATURE SURVEY
    Sr No. Title Author Summery
    Year
    Fraud detection: A realistic Andrea Dal Pozzolo Lack of realism concerns two
    3 modeling & a novel learning et.al main problems:
    strategy. Year: 2018 1. Way & timing in which
    supervised information is
    provided.
    2. Measures used to assess
    fraud detection.
    Applications of whale Mohit verma et.al It copies the pattern of spiral
    4 optimization algorithm Year: 2018 bubblenet hunting pattern of
    humpback.
    It is used to solve numerous
    multidimensional complex
    problems.
    IMPLEMENTATION
    BP NEURAL NETWORK
     The BP algorithm is composed of two parts:
    1. The forward transmission of information.
    Input layer hidden layer output layer
    2. The back propagation of error.
     Hidden layer output is as in formula (1); X1 h1 y1

    output layer output is as in formula (2);


    h2
    y2
    X2
    error function is as in formula (3):

    hn

    y3
    XNA
    THE WHALE OPTIMIZATION
    ALGORITHM
    • The Whale Optimization Algorithm (WOA) is a new heuristic
    optimization algorithm inspired by humpback whale hunting.
    1. Surrounding prey:

    2. Hunting behaviour:

    3. Searching for prey:


    ALGORITHM FLOWCHART
    APPLICATIONS

     Neural Network training


     Economic Dispatch problem
     Workflow Planning of construction site
    ADVANTAGES
     Self-Adaptation

     Self-Organized

     Better fault tolerance

     Robustness

    DISADVANTAGES
     Hard to interpret model.

     Don’t perform well on small data.


    RESULTS
    TRAINING RESULT TEST RESULT
    WOA+BP NOMAL FRAUD DETECTION NOMAL FRAUD DETECTION
    RATE RATE

    A NOMAL 92.75% 96.40%


    34 405 10
    P 1516 93.18% 98.25%

    R FRAUD 33 97.81% 97.83%


    F 111 9 95.44% 8 32 98.04%
    CONCLUSION
    The detection rate of the optimized model woa-bp on the training set is relatively
    close to that of the unoptimized BP model, and the detection rate of the optimized
    model woa-bp on the test set is higher than that of the optimized model. The
    detection rate of the optimized BP model is high, which shows that the woa-bp
    network model has a stronger generalization capability and improves the detection
    accuracy.
    REFERENCE
    1. Phua C, Alahakoon D, Lee V. Minority report in fraud detection:classification of skewed data[J]. Acm Sigkdd
    Explorations Newsletter, 2004, 6(1):50-59.

    2. Soltani N, Akbari M K, Javan M S. A new user-based model for credit card fraud detection based on artificial immune
    system[C]// Csi International Symposium on Artificial Intelligence and Signal Processing. IEEE, 2012:029-033

    3. Renu HCE Sonepat, Suman HCE Sonepat displaystyle. Analysis on Credit Card Fraud Detection Methods //
    International Journal of Computer Trends and Technology (IJCTT) – volume 8 number 1– Feb 2014. ISSN: 2231-2803

    4. Credit Card Fraud Detection: A Hybrid Approach Using Fuzzy Clustering & Neural Network[C]// Second International
    Conference on Advances in Computing and Communication Engineering. IEEE, 2015:494-499.

    5. Mirjalili S, Lewis A. The Whale Optimization Algorithm[J]. Advances in Engineering Software, 2016, 95:51-67.

    6. Sadasivam G S, Subrahmanyam M, Himachalam D, et al. Corporate governance fraud detection from annual reports
    using big data analytics[J]. 2016, 3(1):51

    7. Dal P A, Boracchi G, Caelen O, et al. Credit Card Fraud Detection: A Realistic Modeling and a Novel Learning
    Strategy[J]. IEEE Trans Neural Netw Learn Syst, 2017, PP(99):1-14

    8. M.Rizki A A, Surjandari I, Wayasti R A. Data mining application to detect financial fraud in Indonesia's public
    companies[C]// International Conference on Science in Information Technology. IEEE, 2018:206-211.

    You might also like