Data Analysis Case Study – Auto Claim Assignment

Slides:



Advertisements
Similar presentations
AACEI Contingency Forum Contingency Management
Advertisements

StormingForce.com Motion. StormingForce.com StormingForce’s technology is significantly increasing productivity and quality of manual repetitive tasks.
Quantitative Research and Analytics, Proprietary and Confidential1 Ryan Michaluk
© 2014 Fair Isaac Corporation. Confidential. This presentation is provided for the recipient only and cannot be reproduced or shared without Fair Isaac.
Machine Learning and Data Mining Course Summary. 2 Outline  Data Mining and Society  Discrimination, Privacy, and Security  Hype Curve  Future Directions.
Product Design L5- Ch4: Product Specifications Dr. Husam Arman 1.
1 ACCTG 6910 Building Enterprise & Business Intelligence Systems (e.bis) Data Staging Olivia R. Liu Sheng, Ph.D. Emma Eccles Jones Presidential Chair of.
BORIS MILAŠINOVIĆ FACULTY OF ELECTRICAL ENGINEERING AND COMPUTING UNIVERSITY OF ZAGREB, CROATIA Experiences after three years of teaching “Development.
SOFTWARE PROJECT MANAGEMENT Project Quality Management Dr. Ahmet TÜMAY, PMP.
Software projects Management & Development Alireza Saebi
Michael Solomon Tugboat Software Managing the Software Development Process.
Technical Considerations for Implementing International Credit Projects Jay Tchakarov Director of Product Management HighRadius
QWise software engineering – refactored! Testing, testing A first-look at the new testing capabilities in Visual Studio 2010 Mathias Olausson.
Ihr Logo Data Explorer - A data profiling tool. Your Logo Agenda  Introduction  Existing System  Limitations of Existing System  Proposed Solution.
INFO425: Systems Design INFORMATION X Finalizing Scope (functions/level of automation)  Finalizing scope in terms of functions and level of.
Copyright © 2003, SAS Institute Inc. All rights reserved. Company confidential - for internal use only 1 Know Your Customers SAS® Banking Intelligence.
MODEL BASED GUI TESTING FOR MOBILE APPS 1 Manoj Philip Mathen Specialized Testing Consultant, Infosys.
Business Intelligence Solutions for the Insurance Industry DAT – 13 Data Warehousing Rasool Ahmed.
Lecture #9 Project Quality Management Quality Processes- Quality Assurance and Quality Control Ghazala Amin.
Chapter 6 : Software Metrics
1 Controversial Issues  Data mining (or simple analysis) on people may come with a profile that would raise controversial issues of  Discrimination 
Some Sub-Activities within Requirements Engineering 1.Prototyping 2.Requirements Documentation 3.Requirements Validation 4.Requirements Measurements 5.Requirements.
Property of Progressive Insurance & Casualty Company 1 Agile Test Automation Created and Modified By: Nishant Awasthi January 13, 2009.
Slide 1 Project Management Chapter 4. Slide 2 Objectives ■ Become familiar with estimation. ■ Be able to create a project workplan. ■ Become familiar.
Arben Asllani University of Tennessee at Chattanooga Prescriptive Analytics CHAPTER 8 Marketing Analytics with Linear Programming Business Analytics with.
Software Product Line Material based on slides and chapter by Linda M. Northrop, SEI.
Project quality management. Introduction Project quality management includes the process required to ensure that the project satisfies the needs for which.
Quality Function Deployment Team 6 Rochelle Robles David Harty Johnathan Pesce.
Planning Ahead for Optimal Contact Center Deployment Jim Jenkins
April 29th, Chapter 6 Decision Making 6.1 The Nature of Management decisions 6.1 The Nature of Management decisions 6.2The decision making process.
Do It Strategically with Microsoft Business Intelligence! Bojan Ciric Strategic Consultant
Requirements in the product life cycle Chapter 7.
Yes, Data Management Can Be Agile! Michele Goetz, Principal Analyst.
UBN Software Solutions Pvt. Ltd. Bhubaneswar , Odisha, India ( ) UBN PROJECT CASE STUDY.
Building the Corporate Data Warehouse Pindaro Demertzoglou Data Resource Management.
Capture This! PO105 James Green. Table of Contents Capture Overview Laserfiche Tools Case Scenarios Questions and Answers.
Data Resource Management – MGMT An overview of where we are right now SQL Developer OLAP CUBE 1 Sales Cube Data Warehouse Denormalized Historical.
1 CQA Training QUESTION BANK. 2 CQA Training QUESTION 1 THE TWO DEFINITIONS OF QUALITY ARE: QUALITY MEANS MEETING REQUIREMENTS QUALITY MEANS FIT FOR USE.
QAD Demand Management Patrick Grace, QAD Solutions Consultant September 27th, Fall MWUG Session.
Collage Score Card & Software defect prediction
“Fair settlements come from knowledge and negotiation.”
Cross UG Summit EMEA /8/ :19 PM
Success Stories.
CS15A012 Microsoft Dynamics Lifecycle Services: The key to keeping Microsoft Dynamics AX 2012 implementations running smoothly Murtaza Chowdhury Anupam.
About Client Client is a pioneer in industry that provides catastrophe risk modeling, real-time risk exposure and risk management through available live.
OUTLINE Questions? Comments? Any stories? Questionnaire results
TIM 58 Chapter 3: Requirements Determination
Anjali Yakkundi, Analyst
DEFECT PREDICTION : USING MACHINE LEARNING
Establishing A Data Management Fabric For Grid Modernization At Exelon
PDCA Problem Solving Guide
HCI in the software process
Operationalize your data lake Accelerate business insight
Continuous Performance Engineering
Dr. Morgan C. Wang Department of Statistics
Evaluating Transaction System Performance
12/8/2018 OFC-B270 Why Adoption Matters: Key Factors in Maximizing ROI and Customer Satisfaction in Your Lync Deployment Marc Sanders © 2014 Microsoft.
Jasper Hillebrand Emerging Technologies Think Big Analytics / Teradata
HCI in the software process
Saravana Kumar CEO/Founder - Kovai Atomic Scope – Product Update.
Luxfer Gas Cylinders – S&OP Process Improvement
Luxfer Gas Cylinders – S&OP Process Improvement
HCI in the software process
Megaputer Intelligence
Human Computer Interaction Lecture 14 HCI in Software Process
Service management system at cloud
Predicting Loan Defaults
Project Reflection Implement New Open-System Rating Engine
Copyright © JanBask Training. All rights reserved Get Started with Hadoop Hive HiveQL Languages.
Driving Successful Projects
Presentation transcript:

Data Analysis Case Study – Auto Claim Assignment Ming Sun, American Family Insurance

About Myself 2014 - present 1999 -2005 2005 -2014 Application Development J2EE Web App Java Batch Processing Solution Architecture Big Data Analytics Mobile APP Application Integrations Data Warehouse Integrations Data Science Engineering Repeatable Data Science Pipelines Exploratory Data Analysis Data Lake Design Technology Incubation

Analytical Solution Life-cycle Start Here Current State Bottomline CBA Topline Benefits Data Sources Containerization CI/CD Monitor Pipelines Model Registry Solution Deployment Problem Definition Model Techniques Model Performance Model Pipelines Data Domains Data Quality Data Design Data Blend Data Pipelines Model Development Data Preparation

Problem Definition Scope – Determine if a damaged vehicle should be totaled or repaired at the early stage of auto claims Current State Point Based Model Accuracy < 80% Bottom Line CBA Annual savings amount 10% lift ≈ $500k-$2M Top Line Benefits Impact to customer satisfaction

Problem Definition – Data Sources 3rd normal form DB Claim System – Old (DB2) Partial Data Claims Data Warehouse (DB2) Claim System – New (Oracle) Partial Data No Data 3rd Party Data (daily files)

Data Preparation – Data Domains Handling Assignment (6 - 8 table) 3rd Party Loss Estimates (5 files) Initial Claim (7 - 10 table) Customer Satisfaction (2 files) Code Description 10+ Table Total Loss Workflow (2 - 4 table) Salvage Info (2 table)

Data Preparation – Grain/Quality/Blend The grain of blended dataset - Vehicle Current snapshot of all closed auto collision claims Identify keys to blend claims, 3rd party estimates, and customer satisfaction Profile the blended dataset: record counts, missing values, column value distribution, correlation, etc. This is where the 60% project time is spent

Problem Definition Analysis Current Process: Vehicle Questionnaire Number of questions: 17  12 Questions not answered > 80% Assignment Accuracy ≈ 80% Assigned Repairable, actual Total Loss ≈ 2x % Assigned Total Loss, actual Repairable ≈ x % Mis-assigned Claim Costs Assigned Repairable, actual Total Loss ≈ $ 3y per claim Assigned Total Loss, actual Repairable ≈ $ y per claim

Customer Satisfaction Impact Analysis 5 satisfaction score buckets with 5 being the most satisfied False Positives have the worst impact, followed by False Negatives Customers are happy with True Negatives

Model Development Winner – Logistic Regression Models Misclassification Rate ROC Random Forest 0.136 0.90 Logistic Regression 0.145 0.89 Comparison Category Which Model is Better Technical Performance Random Forest Implementation Cost Logistic Regression (200 vs 1000 hours) Annual Saving Forecast tie The Random Forest model out performs the Points Model from a model performance standpoint. The forecasted annuals savings from these two models are very similar. The time to integrate the Points Model into the claim system is much shorter than the Random Forest model Winner – Logistic Regression

Model Development – Cont’d low Scores Repairable Cutoff Point Manual Review Total Loss Cutoff Point high

Solution Deployment Simplified Vehicle Questionnaire Questions: 12  8 Logistic Regression Points Assignment Claim System - New UI Got rid of the questions that cannot be answered easily. Simplified Vehicle Questionnaire Questions: 12  8 Answers: Y/N  List of Choices

Takeaways Data analysis is critical throughout Keep the data scope reasonable Deep knowledge of business process and data Ease of implementation over model techniques Be conservative when estimating savings Pilot the solution first for 3-6 months to test It is a team effort (analysts, engineers, scientists)

Parting Thought – Data Preparation Most time consuming work Tedious and not glamourous Foundational work – Data Lake Venerable of being the scapegoat