Sharing data and models in software engineering / Tim Menzies [and four others] ; designer, Mark Rogers.

Format:

Book

Author/Creator:

Menzies, Tim, author.

Contributor:

Rogers, Mark, designer.

Language:

English

Subjects (All):

Software engineering.

Computer-aided software engineering.

Physical Description:

1 online resource (415 pages) : illustrations (some color), graphs

Edition:

First edition.

Place of Publication:

Waltham, Massachusetts : Morgan Kaufmann, 2015.

Language Note:

English

System Details:

text file

Summary:

Data Science for Software Engineering: Sharing Data and Models presents guidance and procedures for reusing data and models between projects to produce results that are useful and relevant. Starting with a background section of practical lessons and warnings for beginner data scientists for software engineering, this edited volume proceeds to identify critical questions of contemporary software engineering related to data and models. Learn how to adapt data from other organizations to local problems, mine privatized data, prune spurious information, simplify complex results, how to update models for new platforms, and more. Chapters share largely applicable experimental results discussed with the blend of practitioner focused domain expertise, with commentary that highlights the methods that are most useful, and applicable to the widest range of projects. Each chapter is written by a prominent expert and offers a state-of-the-art solution to an identified problem facing data scientists in software engineering. Throughout, the editors share best practices collected from their experience training software engineering students and practitioners to master data science, and highlight the methods that are most useful, and applicable to the widest range of projects. Shares the specific experience of leading researchers and techniques developed to handle data problems in the realm of software engineering Explains how to start a project of data science for software engineering as well as how to identify and avoid likely pitfalls Provides a wide range of useful qualitative and quantitative principles ranging from very simple to cutting edge research Addresses current challenges with software engineering data such as lack of local data, access issues due to data privacy, increasing data quality via cleaning of spurious chunks in data

Contents:

Front Cover

Sharing Data and Models in Software Engineering

Copyright

Why this book?

Foreword

Contents

List of Figures

Chapter 1: Introduction

1.1 Why Read This Book?

1.2 What Do We Mean by ``Sharing''?

1.2.1 Sharing Insights

1.2.2 Sharing Models

1.2.3 Sharing Data

1.2.4 Sharing Analysis Methods

1.2.5 Types of Sharing

1.2.6 Challenges with Sharing

1.2.7 How to Share

1.3 What? (Our Executive Summary)

1.3.1 An Overview

1.3.2 More Details

1.4 How to Read This Book

1.4.1 Data Analysis Patterns

1.5 But What About …? (What Is Not in This Book)

1.5.1 What About ``Big Data''?

1.5.2 What About Related Work?

1.5.3 Why All the Defect Prediction and Effort Estimation?

1.6 Who? (About the Authors)

1.7 Who Else? (Acknowledgments)

Part I: Data Mining for Managers

Chapter 2: Rules for Managers

2.1 The Inductive Engineering Manifesto

2.2 More Rules

Chapter 3: Rule #1: Talk to the Users

3.1 Users Biases

3.2 Data Mining Biases

3.3 Can We Avoid Bias?

3.4 Managing Biases

3.5 Summary

Chapter 4: Rule #2: Know the Domain

4.1 Cautionary Tale #1: ``Discovering'' Random Noise

4.2 Cautionary Tale #2: Jumping at Shadows

4.3 Cautionary Tale #3: It Pays to Ask

4.4 Summary

Chapter 5: Rule #3: Suspect Your Data

5.1 Controlling Data Collection

5.2 Problems with Controlled Data Collection

5.3 Rinse (and Prune) Before Use

5.3.1 Row Pruning

5.3.2 Column Pruning

5.4 On the Value of Pruning

5.5 Summary

Chapter 6: Rule #4: Data Science Is Cyclic

6.1 The Knowledge Discovery Cycle

6.2 Evolving Cyclic Development

6.2.1 Scouting

6.2.2 Surveying

6.2.3 Building

6.2.4 Effort

6.3 Summary

Part II: Data Mining: A Technical Tutorial

Chapter 7: Data Mining and SE

7.1 Some Definitions

7.2 Some Application Areas.

Chapter 8: Defect Prediction

8.1 Defect Detection Economics

8.2 Static Code Defect Prediction

8.2.1 Easy to Use

8.2.2 Widely Used

8.2.3 Useful

Chapter 9: Effort Estimation

9.1 The Estimation Problem

9.2 How to Make Estimates

9.2.1 Expert-Based Estimation

9.2.2 Model-Based Estimation

9.2.3 Hybrid Methods

Chapter 10: Data Mining (Under the Hood)

10.1 Data Carving

10.2 About the Data

10.3 Cohen Pruning

10.4 Discretization

10.4.1 Other Discretization Methods

10.5 Column Pruning

10.6 Row Pruning

10.7 Cluster Pruning

10.7.1 Advantages of Prototypes

10.7.2 Advantages of Clustering

10.8 Contrast Pruning

10.9 Goal Pruning

10.10 Extensions for Continuous Classes

10.10.1 How RTs Work

10.10.2 Creating Splits for Categorical Input Features

10.10.3 Splits on Numeric Input Features

10.10.4 Termination Condition and Predictions

10.10.5 Potential Advantages of RTs for Software Effort Estimation

10.10.6 Predictions for Multiple Numeric Goals

Part III: Sharing Data

Chapter 11: Sharing Data: Challenges and Methods

11.1 Houston, We Have a Problem

11.2 Good News, Everyone

Chapter 12: Learning Contexts

12.1 Background

12.2 Manual Methods for Contextualization

12.3 Automatic Methods

12.4 Other Motivation to Find Contexts

12.4.1 Variance Reduction

12.4.2 Anomaly Detection

12.4.3 Certification Envelopes

12.4.4 Incremental Learning

12.4.5 Compression

12.4.6 Optimization

12.5 How to Find Local Regions

12.5.1 License

12.5.2 Installing CHUNK

12.5.3 Testing Your Installation

12.5.4 Applying CHUNK to Other Models

12.6 Inside CHUNK

12.6.1 Roadmap to Functions

12.6.2 Distance Calculations

12.6.2.1 Normalize

12.6.2.2 SquaredDifference

12.6.3 Dividing the Data

12.6.3.1 FastDiv

12.6.3.2 TwoDistantPoints.

12.6.3.3 Settings

12.6.3.4 Chunk (main function)

12.6.4 Support Utilities

12.6.4.1 Some standard tricks

12.6.4.2 Tree iterators

12.6.4.3 Pretty printing

12.7 Putting It all Together

12.7.1 _nasa93

12.8 Using CHUNK

12.9 Closing Remarks

Chapter 13: Cross-Company Learning: Handling the Data Drought

13.1 Motivation

13.2 Setting the Ground for Analyses

13.2.1 Wait … Is This Really CC Data?

13.2.2 Mining the Data

13.2.3 Magic Trick: NN Relevancy Filtering

13.3 Analysis #1: Can CC Data be Useful for an Organization?

13.3.1 Design

13.3.2 Results from Analysis #1

13.3.3 Checking the Analysis #1 Results

13.3.4 Discussion of Analysis #1

13.4 Analysis #2: How to Cleanup CC Data for Local Tuning?

13.4.1 Design

13.4.2 Results

13.4.3 Discussions

13.5 Analysis #3: How Much Local Data Does an Organization Need for a Local Model?

13.5.1 Design

13.5.2 Results from Analysis #3

13.5.3 Checking the Analysis #3 Results

13.5.4 Discussion of Analysis #3

13.6 How Trustworthy Are These Results?

13.7 Are These Useful in Practice or Just Number Crunching?

13.8 What's New on Cross-Learning?

13.8.1 Discussion

13.9 What's the Takeaway?

Chapter 14: Building Smarter Transfer Learners

14.1 What Is Actually the Problem?

14.2 What Do We Know So Far?

14.2.1 Transfer Learning

14.2.2 Transfer Learning and SE

14.2.3 Data Set Shift

14.3 An Example Technology: TEAK

14.4 The Details of the Experiments

14.4.1 Performance Comparison

14.4.2 Performance Measures

14.4.3 Retrieval Tendency

14.5 Results

14.5.1 Performance Comparison

14.5.2 Inspecting Selection Tendencies

14.6 Discussion

14.7 What Are the Takeaways?

Chapter 15: Sharing Less Data (Is a Good Thing)

15.1 Can We Share Less Data?

15.2 Using Less Data

15.3 Why Share Less Data?.

15.3.1 Less Data Is More Reliable

15.3.2 Less Data Is Faster to Discuss

15.3.3 Less Data Is Easier to Process

15.4 How to Find Less Data

15.4.1 Input

15.4.2 Comparisons to Other Learners

15.4.3 Reporting the Results

15.4.4 Discussion of Results

15.5 What's Next?

Chapter 16: How to Keep Your Data Private

16.1 Motivation

16.2 What Is PPDP and Why Is It Important?

16.3 What Is Considered a Breach of Privacy?

16.4 How to Avoid Privacy Breaches?

16.4.1 Generalization and Suppression

16.4.2 Anatomization and Permutation

16.4.3 Perturbation

16.4.4 Output Perturbation

16.5 How Are Privacy-Preserving Algorithms Evaluated?

16.5.1 Privacy Metrics

16.5.2 Modeling the Background Knowledge of an Attacker

16.6 Case Study: Privacy and Cross-Company Defect Prediction

16.6.1 Results and Contributions

16.6.2 Privacy and CCDP

16.6.3 CLIFF

16.6.4 MORPH

16.6.5 Example of CLIFF&amp

MORPH

16.6.6 Evaluation Metrics

16.6.7 Evaluating Utility via Classification

16.6.8 Evaluating Privatization

16.6.8.1 Defining privacy

16.6.9 Experiments

16.6.9.1 Data

16.6.10 Design

16.6.11 Defect Predictors

16.6.12 Query Generator

16.6.13 Benchmark Privacy Algorithms

16.6.14 Experimental Evaluation

16.6.15 Discussion

16.6.16 Related Work: Privacy in SE

16.6.17 Summary

Chapter 17: Compensating for Missing Data

17.1 Background Notes on SEE and Instance Selection

17.1.1 Software Effort Estimation

17.1.2 Instance Selection in SEE

17.2 Data Sets and Performance Measures

17.2.1 Data Sets

17.2.2 Error Measures

17.3 Experimental Conditions

17.3.1 The Algorithms Adopted

17.3.2 Proposed Method: POP1

17.3.3 Experiments

17.4 Results

17.4.1 Results Without Instance Selection

17.4.2 Results with Instance Selection

17.5 Summary.

Chapter 18: Active Learning: Learning More with Less

18.1 How Does the QUICK Algorithm Work?

18.1.1 Getting Rid of Similar Features: Synonym Pruning

18.1.2 Getting Rid of Dissimilar Instances: Outlier Pruning

18.2 Notes on Active Learning

18.3 The Application and Implementation Details of QUICK

18.3.1 Phase 1: Synonym Pruning

18.3.2 Phase 2: Outlier Removal and Estimation

18.3.3 Seeing QUICK in Action with a Toy Example

18.3.3.1 Phase 1: Synonym pruning

18.3.3.2 Phase 2: Outlier removal and estimation

18.4 How the Experiments Are Designed

18.5 Results

18.5.1 Performance

18.5.2 Reduction via Synonym and Outlier Pruning

18.5.3 Comparison of QUICK vs. CART

18.5.4 Detailed Look at the Statistical Analysis

18.5.5 Early Results on Defect Data Sets

18.6 Summary

Part IV: Sharing Models

Chapter 19: Sharing Models: Challenges and Methods

Chapter 20: Ensembles of Learning Machines

20.1 When and Why Ensembles Work

20.1.1 Intuition

20.1.2 Theoretical Foundation

20.2 Bootstrap Aggregating (Bagging)

20.2.1 How Bagging Works

20.2.2 When and Why Bagging Works

20.2.3 Potential Advantages of Bagging for SEE

20.3 Regression Trees (RTs) for Bagging

20.4 Evaluation Framework

20.4.1 Choice of Data Sets and Preprocessing Techniques

20.4.1.1 PROMISE data

20.4.1.2 ISBSG data

20.4.2 Choice of Learning Machines

20.4.3 Choice of Evaluation Methods

20.4.4 Choice of Parameters

20.5 Evaluation of Bagging+RTs in SEE

20.5.1 Friedman Ranking

20.5.2 Approaches Most Often Ranked First or Second in Terms of MAE, MMRE and PRED(25)

20.5.3 Magnitude of Performance Against the Best

20.5.4 Discussion

20.6 Further Understanding of Bagging+RTs in SEE

20.7 Summary

Chapter 21: How to Adapt Models in a Dynamic World

21.1 Cross-Company Data and Questions Tackled.

21.2 Related Work.

Notes:

Bibliographic Level Mode of Issuance: Monograph

Includes bibliographical references and indexes.

ISBN:

9780124173071

0124173071

OCLC:

896901265