DEEP LEARNING MODEL FOR DETECTING BREAST CANCER

DEEP LEARNING MODEL FOR DETECTING BREAST CANCER AT EARLY STAGE

Abstract

Breast cancer is the most common cancer among women but in can occur in both the genders. It is accountable for an appalling number of deaths worldwide. In a particularly  developing country like Nigeria, there is a lack of awareness and facilities mostly in rural areas and high rate of instances of breast cancer that is diagnosed in the last stages. However, the early detection of breast cancer can lead to help increase the odds of survival. Nowadays, with the increasing number of patients, manual analysis of medical images becomes tedious, time consuming and unfeasible.  With  the advancement in the field of machine learning,   it is now possible to create an automated and accurate Computer Aided Diagnosis (CAD) system in order to make the entire process of detecting a malignant tumor more resource efficient and time saving through proper utilization. This paper presents the comparative analysis of different machine learning algorithms and their results in predicting cancerous tumors. The proposed model uses supervised machine learning algorithms such as Random Forest, Support Vector Machine, K-Nearest Neighbors, Naïve Bayes and Logistic Regression with and without PCA on a dataset with 30 features extracted from a digitized image of a fine needle aspirate (FNA) of a breast mass. Deep learning models like Artificial Neural Network and Convolutional Neural Network are used and their performances are compared. From the comparative analysis, it is observed that the deep learning models outperform all other classifiers and achieves impressive scores across multiple performance metrics such as Accuracy of 98.83%, Precision of 98.44% and Recall of 100%.

LISTED CONTENTS FOR DEEP LEARNING MODEL FOR DETECTING BREAST CANCER Pdf

  • CHAPTER ONE
  • Introduction
  • Motivation
  • Objective
  • Thesis Orientation
  • CHAPTER TWO
  • Literature Review
  • Related Works .
  • Machine Learning
  • proposed Model
  • Dataset
  • Data Visualization
  • Histogra
  • Heatmap
  • Data Preprocessing
  • Categorical Variable Conversion
  • Feature Scaling
  • Principal Component Analysis (PCA)
  • Data Reshaping
  • Train-Test Split
  • Neural Network Layers
  • Algorithms
  • Random Forest
  • Support Vector Machine
  • K-Nearest Neighbor
  • Logistic Regression
  • Naïve Bayes
  • Artificial Neural Network (ANN)
  • Convolutional Neural Network (CNN
  • CHAPTER THREE
  • Result Analysis
  • Performance Metrics
  • Confusion matrix
  • Accuracy
  • Precision
  • Recall or Sensitivity
  • F1 Score
  • Receiver Operating Characteristics (ROC) Curve
  • Area under the ROC Curve (AUC)
  • Model Performances
  • Random Forest (RF)
  • Support Vector Machine (SVM)
  • K-Nearest Neighbors (KNN)
  • Logistic Regression (LR)
  • Naïve Bayes
  • Artifical Neural Network (ANN)
  • Convolutional Neural Network (CNN)
  • CHAPTER FIVE
  • Discussion
  • Conclusion
  • References
Download Full Material-N5000

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Post

DESIGN AND IMPLEMENTATION OF A COMPUTERIZED AIRPLANE PASSENGER MANIFEST INFORMATION

DESIGN AND IMPLEMENTATION OF A COMPUTERIZED AIRPLANE PASSENGER MANIFEST INFORMATION  A CASE STUDY OF SOSOLISO AIRWAYS

CHAPTER ONE/INTRODUCTION

1.1     OVERVIEW OF THE EXISTING SYSTEM

This passenger manifest information (PMI) is a kind of records that help individual flight to organized their passenger in a flight. This record in the flight. This record help them to know the members of passenger in the flight, without their records they can known the numbers of passenger in the flight. So the passenger manifest information contained the:

  • Full name
  • Date of birth
  • Social security number
  • Emergency contact
  • Telephone number

From passenger traveling on flight with sosoliso airline (e.g., during the sosoliso airline disaster family assistance act of December 10th 2005. this law was passed to address the difficult associated with notification of the families in the aftermath of sosoliso airline crashes). This proposal is also being issued to fulfill a recommendation contained in the initial and final report of families that has be affect in the crashes. This passenger manifest information (PMI), is used to point out identify data elements  that may point to a passenger who represents a security risk. Once highlighted, analytical specialists can them manually review these records and identify those  passengers who should be subjected to additional screening upon their arrival.

 

  • DESCRIPTION AND ANALYSIS OF THE EXISTING SYSTEM

This study is all about the passengers manifest information (PMI) which is a records of individual passenger in the flight.

The researcher visited sosoliso airline and find out that the passenger manifest information contained the records of passenger (e.g., full name, date of birth and or social security number, emergency contact and telephone number). Which help them to organize their flight.

The examination of the existing system to determine what must be accomplished and the best method of accomplishing the necessary operation is know as SYSTEM ANALYSIS. The stages involved in the system analysis are listed below.

  • The type of processing to be done by the system
  • The input data “captured’ for the system
  • The output required
  • The file which could be kept and used
  • The nature of processing i.e. its logic procedure stage and frequency.

 

  • METHOD OF DATA COLLECTION

The researcher visited sosoliso airline and find out that the passenger manifest information (PMI) contained the records of passengers, which help them to organize their flight.

SOURCE OF DATA

This data is used in this research work were collected from primary and secondary source hence, they comprise of the following sources:

  • Direct interview in (sosoliso airline)
  • The internet
  • Unpublished books (project)
  • Pamphlets of general knowledge.

1.4     INTERVIEW METHOD

The researcher use the method of unstructured interview to carried out the research to know about the passenger manifest information in sosoliso airline. It ash one of the staff in the engineer department, the researcher know that the passenger manifest information contain the name, date of birth, contact address etc, of a passengers that booking for ticket to travel in a flight in the sosoliso airline company.

 

REFERENCE TO WRITING TEXT

The researcher used this material to carried out it’s project work. This material are:

  • Sosoliso airline company (engineer department)
  • Unpublished books (project)
  • From the federal register: March 13 1997 volume 62, number 49. online via GPO access (wais.access.gpo.gov)
  • Presented by the IATA: Agenda items: facilitation and security of travel documents and boarder control 2.4: advance passenger information (API) airline reservation system and passenger name record (PMI) access by states.

 

INPUT ANALYSIS

The necessary input to the system includes the following:-

  • Full name
  • Date of birth
  • Social security number
  • Emergency contact
  • Telephone number

The full name of individual passenger that booking for the flight indicates the records of passenger that travel in sosoliso airline.

The date of birth. Is the day that the passenger was born in that particular city or country.

The social security number: is a kind of information that no two person have it belong to one person alone.

Emergency contact: Is the places of the passengers that booking for the flight in Sosoliso airline.

Telephone number: This is a kind of information that can be used to reach the passenger that booking in sosoliso airline of this is globalization network that has be used to reach people in different places. Below is a diagrammatic representation of the passenger manifest information in Sosoliso airline Enugu”.

Where are you going?

I’m flying: from To

Options: class

Return one way

Class: Economy first class

WHERE ARE YOU GOING

Leave: Day month morning coming back: Day month evenining number of passengers:

Adults (over 12 years) 1 children aged 2 – 12 years 0 infants under 2 years 0

TITLE SELECT A TITLE

First name:

Surname:

Children name (if any):

PROCESS ANALYSIS

The process involve the transformation of available input data inot useful output in this particular study, this entails matching projects with their appropriate passenger manifest information in such a way as to satisfy the criteria of requirements and suitability: Thus, a project is picked from the scheduled proposed project list, its passenger information requirements matched against the passenger manifest information attributes of various location until a perfect already obtained. If a similar project already exists in the location that matched another and they may be good match is used a s a substitute site for the project.

OUTPUT ANALYSIS

After processing using the necessary input data, the output from the system include: the passenger information such has, full name, date of birth emergency contact, social security number, and telephone number.

PROBLEMS OF THE EXISTING SYSTEM

The existing system has numerous disadvantages and inherent dangers.

  • Basic Approach: Passengers would be defined broadly to include confirmed, ticketed passengers as well as standing bys, walk-ups, lap infants, those rerouted from another flight or our carrier, and non-revenue passengers. At this time, we expect that the sosoliso passenger manifest information would consist of passenger:
    1. Full name;
    2. Date of birth (DOB)
  • Social security number;
  1. Contact name
  2. Contact telephone number.

Further, we envision the information would be transmitted to the department of transportation (DOT) and the national transportation safety board (NTSB) as soon as possible, but no later than three hours, after the sosoliso disaster.

  • Frequent flyer information and a sosoliso passenger manifest information requirement. We understand that more extensive passenger information is kept on hand for frequent flyers, and that frequent flyers account for over one-half of all passengers traveling in sosoliso airline.
  • Privacy considerations and fraud issues. What privacy issued are raised by a sosoliso passenger manifest information requirement as outlined above? Will manifest information be subject to subpoena by private litigants and law enforcement agencies? What fraud issued, if any, are raised by implementing the above sosoliso passenger manifest information requirement? What are the implications for personal privacy that would result if sosoliso airline were required to collect any of the following information from passengers: full name, date of birth, social security number, emergency contact and phone number? What types of safeguards, if any, should be placed upon the passenger manifest information that is collected by sosoliso airline.
  • Coverage of sosoliso passenger manifest information requirement and the differing implications, if any, for different types of airline that would be covered. We envision and charter operators would be covered by a sosoliso passenger manifest information requirement”, scheduled and charter airline, as well as air trails and commuter.

 

JUSTIFICATION OF THE NEW SYSTEM

(1)     It satisfies the user requirement at the right time with in an acceptable level of accuracy

(2)     It provides an effective security measure in order to avoid loss of stored data in the master file and also avoid fraud.

(3)     It provides an adequate coding system to aid identification, comparison sorting, unification and elimination of ambiguities (error).

(4)     It provides check and control, which are capable of detecting and dealing with exceptional circumstances and errors.

(5)        It provides an adequate insight on policy matter and their effects on the passenger manifest information requirement systemDownload Full Material-N5000

COURSE REGISTRATION AND RESULT PROCESSING SYSTEM IN COMPUTER SCIENCE DEPARTMENT

COURSE REGISTRATION AND RESULT PROCESSING SYSTEM IN COMPUTER SCIENCE DEPARTMENT UNIVERSITY OF NIGERIA, NSUKKA

ABSTRACT

Course Registration and Result processing system form the key activities in the life span of a student. The demand for effective and efficient result computation and output lead to the production of the automated result process. The digitalized process provides the capabilities of centralized repository for storage, management and dissemination of result information to those concerned. A study into the manual process reveals the inefficiencies and the rigorous nature of course registration and result processing, therefore the existing manual process was analyzed and converted into an automated system. The manual activities was modeled with an object-oriented methodology and the system was implemented with PHP(hypertext processor), a scripting language and MYSQL database management systems. It provides a robust database that generates various reports that is relevant to the department. The result of this research work is a system that automates the course registration and examination result processing for the Department of Computer Science, which can be used in any department of most universities in Nigeria.

TABLE OF CONTENTS
LIST OF ITEMS PAGE NUMBER
Title Page …………………………………………………………. i
Certification page …………………………………………………. ii
Approval Page ..…………………………………………………… iii
Dedication …………………………………………………………. iv
Acknowledgement Page …………………………………………… v
Abstract …………………………………………………………….. vi
Table of Contents.…………………………………………………… vii
List of Figures……………………………………………………….. ix
List of Tables………………………………………………………… xi
Chapter 1: Introduction ………………………………………… 1
1.0 Introduction …………………………………………….. 1
1.1 Statement of Problem .………………………………….. . 3
1.2 Objectives ……………………………………………… . 4
1.3 Scope of the Study ……………………………………….. 4
1.4 Significance of the Study…………………………………. 4
Chapter 2: Literature Review ………………………………………………… 5
2.0 Introduction……………………………………………….. 5
2.1 Theoretical Background …………………………………… 5
2.2 Review of Related Literature………………………………. 10

Chapter 3: Systems Analysis and Design…………………………. 21
3.0 Introduction ……………………………………………… 21
3.1 Analysis of the Existing System………………………….. 22
3.2 Description of the Proposed System……………………… 25
3.3 Design of the Proposed System…………………………… 27
3.3.1 Design Methodology……………………………… 27
3.3.2 Justification for the use of the design methodology. 27
3.3.3 Design tool………………………………………… 28
3.3.4 Database design…………………………………… 32
3.3.5 Input and output design…………………………… 35
3.4 Systems Architecture………………………………………. 45
Chapter 4: System Implementation………………………………… 46
4.0 Introduction………………………………………………… 46
4.1 Development Environment…………………………………. 46
4.2 Choice of Programming Language used……………………. 47
4.3 Implementations Architecture………………………………. 47
4.4 Software Testing……………………………………………. 48
4.5 Documentation……………………………………………… 49
4.5.1 Software Modules………………………………49
4.5.2 User Guide/Manual…………………………… 52
4.5.3 Software Installation………………………… 54
4.5.4 Source Code Listing…………………………….54
Chapter 5: Summary and Conclusion………………………………. 55
5.0 Summary……………………………………………………. 55
5.1 Conclusion………………………………………………….. 56
5.2 Recommendations………………………………………….. 56
References
Appendices

INTRODUCTION

  • INTRODUCTION

Automated course registration and exam processing system is a software that facilitates the students course registration and processing of examination result. It will build in an automated information system capability that will ease the stress and enhance all the activities involved in course registration and result processing. The occurrence of examination and assessment irregularities can seriously damage public confidence in the validity and legitimacy of examination and assessment results and should be dealt with as a matter of urgency within the judicial framework established for this purpose[12]. There are different computer programs in different tertiary institutions for computing examination results. However, beyond examination result computation, not many programs in use provide multi-level aggregated data of student population and academic progress at various stages of studentship. The need for data use to inform administrative decisions in tertiary institutions have been emphasized for the purpose of better planning[1]. Almost every section of the educational system requires information processing. With the deployment of a good information system, wear and tear occasioned during data retrieval and handling, and sometimes loss is greatly minimized. Data sharing is less difficult and reproduction becomes a lot more cheaper. Furthermore, information system when employed to the educational sector, prompts the following:- prompt access to students’ personal and course information, instant student information updating, automatic computation of the Grade Point Average (GPA), monitoring of failed courses, quick production of result, storing course information such as course code, course description, course unit and scores for the purpose of GPA computation, and producing user-friendly data entry screens for ease of use. It is unfortunate that most universities in Nigeria still adapt the manual system of course registration and examination result processing. The errors associated with the existing manual method of results computation make it not only desirable but imperative that automated approach be used to the full in measuring students’ progress. The manual methods being employed suffer a number of set-backs. They make the process to be time-consuming and prone to error. They lead to examination results being published late, sometimes with wrong grades being entered and students’ grade point averages being wrongly computed. They also lead to loss or misplacement of student’s result. In some cases, it leads to incoherent course registration. However some departments in some universities in Nigeria are gradually advancing with the new information system processes. Taking the University of Nigeria, Nsukka as a case study,

The current course registration in the university of Nigeria, Nsukka takes the following format. Students on resumption renew their registration and register for courses every new semester. Before the current system was implemented, students register their courses with the different lecturers handling the courses. The list is compiled and sent to the appropriate centre. Later the system improved, a better system came into place where students go online and upon entering their registration number and department, the list of available courses for a particular level for a semester is shown and they register for them by selecting the appropriate box against their chosen courses. This is then submitted online and a printed copy is taken to the appropriate centre for approval. Examination result processing draws all the registered courses together and applies a criteria to process each student’s result at different semesters, session and at the end of the school programme. But the system currently in use poses challenges to the exam officers in the departments. It is observed that students do not register courses as required by the department. This could either be because the students deliberately registered wrongly or the system cannot enlist all the semester courses at a time. Secondly, the current system is unable to enlist all failed courses that were taken in the previous semesters and does not compel students to register them before proceeding to the new ones. It does not have the capability to efficiently reconcile all registered courses with their respective result. Therefore this project work is targeted to eliminate most of the problems inherent in the current system. It would provide a set of structures that keeps databases of all registered courses at all levels enlisting the core courses, electives and required ancilliary courses, it keeps student’s personal information. It will also keep past record of student’s result and a database of all components required for processing results. It will have component that extract each student’s result corresponding to the registered courses and computes the required result at any level and time.

1.1      STATEMENT OF THE PROBLEM

The problems identified in the old system are as follows:

  1. The old system does not have a unified database system.
  2. There is no structure that compels students to register all outstanding courses before proceeding to the recent ones.
  3. Student’s test and exam scores may be entered wrongly when manually done.
  4. It is very cumbersome to manually reconcile all registered courses from different files sources into their work sheet for the purpose of result computation.

1.2      AIM AND OBJECTIVES

The main aim of this project work is to provide a centralized database system that prompts correct course registration and keep all result information of a student.

The objectives to realize the aim of this project work are as follows:

  1. Enforce a system that would automatically pool fail courses into the student’s registration portal and allow them to register them before they can register new ones.
  2. Put in place a unified database system that coordinates course registration and exam processing.
  3. Provide a mechanism of transmitting course registration and examintion results to all stakeholders especially Department, Exams and Registry.
  4. Produce all kinds of results expected from the university.

1.3      SCOPE OF THE STUDY  

The new system designed is meant to handle course registration and exam result processing for the Computer Science department in UNN only.

1.4 SIGNIFICANCE OF THE STUDY

Students stand to greatly benefit from this project work. The new system will enable students to see all failed courses and give them the opportunity to register and rewrite them on time. There would be reduced case of overstay.

Result computation would be easier and convenient for the exam officers as all result information will be kept and generated from the system.

The new system would enhance the result processing performance as it will reduce delay in computing student’s result.Download Full Material-N5000

SUBSCRIPTION FRAUD DETECTION SYSTEM

SUBSCRIPTION FRAUD DETECTION SYSTEM

ABSTRACT

This project is concerned with GSM subscription fraud detection system using artificial neural network technique. Fraud is a multi-billion problem around the globe with huge loss of revenue. Fraud can affect the credibility and performance of telecommunication companies. The most difficult problem that faces the industry is the fact that fraud is dynamic, which  means that whenever fraudsters feel that they will be detected, they device other ways to circumvent security measures. In such cases, the perpetrators intention is to completely avoid or at least reduce the charges for using the services. Subscription fraud is one of the major types of telecommunication fraud in which a customer obtain an account without intention to pay the bill. Thus at the level of a phone number, all transactions from the number will be fraudulent. In such cases abnormal usage occurs throughout the active period of the account; which is usually used for call selling or intensive self usage. This provides a means for illegal high profit business for fraudsters requiring minimal investment and relatively low risk of getting caught. A system to prevent subscription fraud in GSM telecommunications with high impact on long distance carriers is proposed to detect fraud. The system employs adaptive flexible techniques using advanced data analysis like Artificial Neural Networks (ANN). Fed with raw data, a neural network can quickly learn to pick up patterns of unusual variations that may suggest instances of fraud on a particular account. A total of 158 data samples were collected, trained and tested using a model that allows identifying potential fraudulent customers at the time of subscription. The result shows that 80% of the prediction accuracy has been obtained. From the result produced, artificial neural network has a potential to be used for detecting subscription fraud in telecommunication.
Download Full Material-N5000