Actuarial Model Optimal Fit

1

• Actuarial Model Optimal Fit
• for Fitting Data to
• Probability Density Functions (PDFs)

2
Probably Dense

• Team Leader Daniel Hanson
• Lead Developer Matthew Miller
• Q.A. Manager Jon Swenson
• Lead Tester Edward Badgley
• Advisor Dr. James Schwing
• Client Dr. Yvonne Chueh

3
Problem

• Actuarys often need to find the best fit PDF for
  a set of data.
• Many programs fit data to one PDF, but not the
  mixed distributions often encountered in the
  Actuarial Field.

4
Solution

• We will develop a program to find a best fit
  distribution for a data set.
• The program will find a best fit mixed
  distribution if required.

5
Overview

• Project Requirements
• Prototype Demonstration
• Quality Assurance Plan

6
Project Requirements

• Matthew Miller
• Lead Developer

7
Computer Requirements

• Development
• IBM PC compatible Pentium equivalent or higher
• 128MB RAM or greater
• Windows 2000 or XP
• C/C compiler
• Target Installation
• IBM PC compatible Pentium equivalent or higher
• 64MB RAM, recommended 128MB or greater
• Windows 95/98, NT 4.0, 2000, or XP

8
Functional Requirements

• Run Calculations
• Data Statistics
• Maximize the Maximum Likelihood Estimate (MLE)
  for the 22 PDFs and if needed, the mixed
  distributions
• Determine the Best Fit PDF
• Request Plots
• Histogram of data
• Best fit PDF
• Best fit PDF with Histogram of data

9
Functional Requirements (contd)

• Enter Data
• Data is entered or imported
• Open Workspace
• Data is read in
• Variables are initialized
• Save Workspace
• Data is written to file
• Statistics are written to file

10
Non-Functional Requirements

• Reliability
• Program Execution Time
• Calculations finish within 10 minutes
• Ease of Use
• Feel of a standard office application
• Development Deadline
• Mid-March
• Quality Assurance Conformance

11
Prototype Demonstration

• Edward Badgley
• Lead Tester

12
Quality Assurance Plan

• Jon Swenson
• Q.A. Manager

13
Q.A. Overview

• In order to maximize team efficiency and minimize
  the risk of errors encountered during this
  project, QA standards and procedures have been
  developed specific to our project.
• Document Standards
• Coding Standards
• User Interface Guidelines
• Change Control Process
• Testing Process
• Risks and Risk Management

14
Document Standards

• Provides a framework to allow consistency of
  documentation throughout the project and to
  ensure the completeness of project documentation.
• All deliverable documents use the Professional
  Report template found in MS Word.
• The creation of documentation shall be
  accomplished by using a table of contents
  consisting of sections which satisfy document
  requirements specific to each deliverable
  document assignment.
• All deliverable documents must have rough drafts
  completed at least three days prior to delivery
  for group review and editing.

15
Coding Standards

• Provides a framework to allow consistency of code
  generated throughout the project and to ensure
  the delivery of well-written code.
• Module headers will describe the flow and
  operation of the module
• Function headers will describe the flow and
  operation of the function
• One code statement per line

16
User Interface Style Guidelines

• Provides a framework to allow consistency of GUIs
  generated throughout the project and to ensure
  the delivery of well-designed, user-friendly
  graphical user interfaces.
• All GUIs will be designed to mimic the look and
  feel of a Microsoft Windows GUI.

17
Change Control Process

• Provides a framework to allow all parties
  affected by proposed changes to accepted work
  products or specifications to assess the
  schedule, and/or product impact of the changes.
• Formative Development
• Product Acceptance
• Change Proposal
• Approving or Rejecting Proposed Changes
• Change Control Board

18
Testing Process

• Test each module at completion
• White Box Testing
• Test integrated program upon completion of all
  modules
• White Box Testing
• Black Box Testing

19
Risks and Risk Management

• Time Constraints
• Probability High
• Impact Serious
• Team Experience
• Probability Moderate
• Impact Tolerable
• Loss of Team Member
• Probability Low
• Impact Catastrophic

• Poor Usability
• Probability Low
• Impact Serious
• User Non-acceptance
• Probability Low
• Impact Serious
• Platform Issues
• Probability Low
• Impact Tolerable

20
Status Report

• Developing the Algorithm for the calculations.
• Writing the Software Requirements document.
• Creating the Quality Assurance document.

21
Conclusion

• Challenging Development Process
• Software Requirements
• Prototype
• Quality Assurance Plan