-----------------

[Wink]

1. Introduction
2. Background
   2.1 Felligi-Sunter Model of Record Linkage
   2.2 Empirical Data Bases and Description of Matching Procedures
   2.3 Expectation-Maximization Algorithm
   2.4 Frequency-Based Weighting
   2.5 String Comparator Metrics
   2.6 Linear Sum Assignment Algorithm
3. Results
4. Discussion
   4.1 How the Quality of Files Affects Matching Results
   4.2 String Comparison and Weight Adjustment in the Fellegi-Sunter
   Model
   4.3 Weights Produced by the EM Algorithm
       4.3.1 Weights From General Statistical Fitting Procedures
       4.3.2 Comparison With Existing Weighting Methods
   4.4 Frequency-Based Weights
   4.5 Linear Sum Assignment Procedure
   4.6 Determination of Cutoff Weights
5. Summary and Conclusions
   5.1 Summary
   5.2 Conclusion
   5.3 Future Work

--------------------

[Wick]

1. Introduction
2. Motivation
3. Usage
   3.1 Input
   3.2 Output
4. Helpers
5. Strategy
   5.1 Case Study: Baseball
   5.2 Case Study: Ozone
   5.3 Other Issues
6. Related Work
7. Conclusion

--------------------

[Efr]

1. Introduction
2. Bootstrap methods
3. Estimating the median
4. Error rate estimation in discriminant analysis
5. Relationship with the jackknife
6. Wilcoxon's Statsitic
7. Regression models
8. Remarks

--------------------

[Lei]

1. Introduction
   1.1 Prediction Sets and Density Level Sets
   1.2 Main Results
   1.3 Related Work
2. Conformal Prediction
3. Conformal Prediction With Kernel Density
   3.1 The Method
   3.2 An Approximation
   3.3 Asymptotic Properties
4. Choosing The Bandwidth
5. Numerical Examples
   5.1 2D Gaussian Mixture
   5.2 Further Simulations
       5.2.1 Double Exponential Distribution
       5.2.2 Three-Dimensional Data
   5.3 Application to Breast Cancer Data
(Appendices)

-------------------

[Mick]

[unlabelled introduction]
ROC Analysis in the Medical Literature
Confidence and Accuracy in Eyewitness Identification
How to Construct an Eyewitness Memory ROC
The Problem With Comparing Single HR-FAR Pair
Experiment 1
  Method
    Participants
    Materials
    Procedure
  Results
Experiment 2
  Method
    Participants
    Materials
    Procedure
  Results
Results and Discussion
  The Diagnosticity Ratio does not identify the more accurate
    lineup procedure
  The Diagnosticity Ratio does not identify the optimal HR
    versus FAR trade off
General Discussion

---------------