Tutorial on logistic-regression calibration and fusion:converting a score to a likelihood ratio

Australian Journal of Forensic Sciences

Volumn 45, Issue 2, 2013, Pages 173-197

  Morrison, Geoffrey Stewart ^a  

^a UNIVERSITY OF NEW SOUTH WALES   (Australia)

Author keywords

Calibration; Forensic science; Fusion; Likelihood ratio; Logistic regression; Score

Indexed keywords

ARTICLE; FINGER DERMATOGLYPHICS; FORENSIC SCIENCE; HUMAN; LOGISTIC REGRESSION ANALYSIS; LOGISTIC REGRESSION CALIBRATION; LOGISTIC REGRESSION FUSION; MAXIMUM LIKELIHOOD METHOD; NORMAL DISTRIBUTION; PROBABILITY; SPEECH DISCRIMINATION; STATISTICAL MODEL; VARIANCE;

EID: 84878013044     PISSN: 00450618     EISSN: 1834562X     Source Type: Journal    
DOI: 10.1080/00450618.2012.733025     Document Type: Article

References (40)

1. Ramos-Castro D, González-Rodríguez J, Ortega-García J. Likelihood ratio calibration in a transparent and testable forensic speaker recognition framework. In: Proceedings of the Odyssey Language and Speaker Recognition Workshop. IEEE. 2006. doi: 10. 1109/ODYSSEY. 2006. 248088
2. González-Rodríguez J, Rose P, Ramos D, Toledano DT, Ortega-García J. Emulating DNA - Rigorous quantification of evidential weight in transparent and testable forensic speaker recognition. IEEE Trans Audio Speech Lang Process. 2007; 15: 2104-2115. doi: 10. 1109/TASL. 2007. 902747
3. González-Rodríguez J, Ramos D. Forensic automatic speaker classification in the 'coming paradigm shift'. In: Müller C, editor. Speaker Classification I: Fundamentals, Features, and Methods. Heidelberg (Germany): Springer-Verlag; 2007. p. 205-217. doi: 10. 1007/978- 3-540-74200-5-11. (Pubitemid 350121203)
4. Ramos Castro D. Forensic evaluation of the evidence using automatic speaker recognition systems [PhD dissertation]. Madrid (Spain): Universidad Autónoma de Madrid; 2007.
5. Thiruvaran T, Ambikairajah E, Epps J. FM features for automatic forensic speaker recognition. In: Proceedings of Interspeech 2008 Incorporating SST. International Speech Communication Association; 2008. p. 1497-1500.
6. Morrison GS. Likelihood-ratio forensic voice comparison using parametric representations of the formant trajectories of diphthongs. J Acoust Soc Am. 2009; 125: 2387-2397. doi: 10. 1121/1. 3081384
7. Enzinger E. Formant trajectories in speaker recognition [master's thesis]. Vienna (Austria): Universität Wien; 2009.
8. Morrison GS, Zhang C, Rose P. An empirical estimate of the precision of likelihood ratios from a forensic-voice-comparison system. Forensic Sci Int. 2011; 208: 59-65. doi: 10. 1016/j. forsciint. 2010. 11. 001
9. Morrison GS, Thiruvaran T, Epps J. Estimating the precision of the likelihood-ratio output of a forensic-voice-comparison system. In: Cernocký H, Burget L, editors. Proceedings of Odyssey 2010: The Language and Speaker Recognition Workshop, Brno, Czech Republic. International Speech Communication Association; 2010. p. 63-70.
10. Morrison GS, Thiruvaran T, Epps J. An issue in the calculation of logistic-regression calibration and fusion weights for forensic voice comparison. In: Tabain M, Fletcher J, Grayden D, Hajek J, Butcher A, editors. Proceedings of the 13th Australasian International Conference on Speech Science and Technology, Melbourne, Australia. Australasian Speech Science and Technology Association; 2010. p. 74-77.
11. Morrison GS. A comparison of procedures for the calculation of forensic likelihood ratios from acoustic-phonetic data: Multivariate kernel density (MVKD) versus Gaussian mixture model-universal background model (GMM-UBM). Speech Comm. 2011; 53: 242-256. doi: 10. 1016/j. specom. 2010. 09. 005
12. Morrison GS. Measuring the validity and reliability of forensic likelihood-ratio systems. Sci & Just. 2011; 51: 91-98. doi: 10. 1016/j. scijus. 2011. 03. 002
13. Rose P. Forensic voice comparison with Japanese vowel acoustics - a likelihood ratiobased approach using segmental cepstra. In: Lee W-S, Zee E, editors. Proceedings of the 17th International Congress of Phonetic Sciences, Hong Kong, China. Hong Kong: Organizers of ICPhS XVII at the Department of Chinese, Translation and Linguistics, City University of Hong Kong. 2011. p. 1718-1721.
14. Rose P. Forensic voice comparison with secular shibboleths - a hybrid fused GMMmultivariate likelihood ratio-based approach using alveo-palatal fricative cepstra. In: Proceedings of the International Conference on Speech Signal Processing, Prague, Czech Republic. IEEE. 2011. p. 5900-5903. doi: 10. 1109/ICASSP. 2011. 5947704
15. Zhang C, Morrison GS, Thiruvaran T. Forensic voice comparison using Chinese /iau/. In: Lee W-S, Zee E, editors. Proceedings of the 17th International Congress of Phonetic Sciences, Hong Kong, China. Hong Kong: Organizers of ICPhS XVII at the Department of Chinese, Translation and Linguistics, City University of Hong Kong. 2011. p. 2280- 2283.
16. Neumann C, Evett IW, Skerret J. Quantifying the weight of evidence from a forensic fingerprint comparison: a new paradigm. J Royal Statist Soc A. 2012: 175; 371-415. doi: 10. 1111/j. 1467-985X. 2011. 01027. x
17. Evett IW. Quo vadis, forensic science? Keynote presentation presented at: 6th European Academy of Forensic Science Conference; 2012 August 20-24; The Hague (The Netherlands).
18. Meuwly D, Ramos D, van Leeuwen D, Haned H, Harakin R. Assessment of performance of forensic likelihood-ratio-based evaluation methods. Workshop presented at: 6th European Academy of Forensic Science Conference; 2012 August 20-24; The Hague (The Netherlands).
19. Morrison GS. An introduction to data-based calculation of likelihood ratios and assessment of validity and reliability. Workshop presented at: 6th European Academy of Forensic Science Conference; 2012 August 20-24; The Hague (The Netherlands).
20. Ramos D. Reliable support: Measuring calibration of likelihood ratios. Keynote presentation presented at: 6th European Academy of Forensic Science Conference; 2012 August 20-24; The Hague (The Netherlands).
21. Platt JC. Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. In: Smola AJ, Bartlett P, Schölkopf B, Schuurmans D, editors. Advances in Large Margin Classifiers. Cambridge (MA): MIT Press; 2000. p. 61- 74.
22. Pigeon S, Druyts P, Verlinde P. Applying logistic regression to the fusion of the NIST'99 1-speaker submissions. Digit Signal Process. 2000; 10; 237-248. doi: 10. 1006/dspr. 1999. 0358 (Pubitemid 30592153)
23. Brümmer N, du Preez J. Application independent evaluation of speaker detection. Comput Speech Lang. 2006: 20; 230-275. doi: 10. 1016/j. csl. 2005. 08. 001 (Pubitemid 41787538)
24. van Leeuwen DA, Brümmer, N. An introduction to application- independent evaluation of speaker recognition systems. In: Müller C, editor. Speaker Classification I: Fundamentals, Features, and Methods. Heidelberg, Germany: Springer-Verlag; 2007. p. 330-353. doi: 10. 1007/978-3-540-74200-5-19 (Pubitemid 350123449)
25. Brümmer N, Burget L, Cernocký JH, Glembek O, Grézl F, KarafiátM, van Leeuwen DA, Matejka P, Schwarz P, Strasheim A. Fusion of heterogenous speaker recognition systems in the STBU submission for the NIST SRE 2006. IEEE Trans Audio Speech Lang Process. 2007: 15: 2072-2084. doi: 10. 1109/TASL. 2007. 902870
26. Brümmer N. Measuring, refining and calibrating speaker and language information extracted from speech [PhD dissertation]. Stellenbosch (South Africa): Stellenbosch University; 2010.
27. Morrison GS. Forensic voice comparison. In: Freckelton I, Selby H, editors. Expert evidence. Sydney (Australia): Thomson Reuters; 2010. ch. 99.
28. Reynolds DA, Quatieri TF, Dunn RB. Speaker verification using adapted Gaussian mixture models. Digit Signal Process. 2000; 10: 19-41. doi: 10. 1006/dspr. 1999. 0361 (Pubitemid 30592166)
29. Agresti A. An introduction to categorical data analysis. 2nd ed. Hoboken (NJ): Wiley; 2007.
30. Hastie T, Tibshirani R, Friedman J. The elements of statistical learning: Data mining, inference, and prediction. 2nd ed. New York: Springer; 2009.
31. Hosmer DW, Lemeshow S. Applied logistic regression. 2nd ed. New York: Wiley; 2000.
32. Menard S. Applied logistic regression. Thousand Oaks (CA): Sage; 2002.
33. Morrison GS. Logistic regression modelling for first- and second-language perception data. In: Solé MJ, Prieto P, Mascaró J, editors. Segmental and prosodic issues in Romance phonology. Amsterdam: John Benjamins; 2007. p. 219-236.
34. Pampel FC. Logistic regression: A primer. Thousand Oaks (CA): Sage; 2000.
35. Morrison GS, Kinoshita Y. Automatic-type calibration of traditionally derived likelihood ratios: Forensic analysis of Australian English /o/formant trajectories. In: Proceeding of Interspeech 2008 Incorporating SST 2008. International Speech Communication Association; 2008. p. 1501-1504.
36. Aitken CGG, Lucy D. Evaluation of trace evidence in the form of multivariate data. Appl Statist. 2004; 53: 109-122. doi: 10. 1046/j. 0035-9254. 2003. 05271. x (Pubitemid 38254794)
37. Aitken CGG, Lucy D. Corrigendum: Evaluation of trace evidence in the form of multivariate data. Appl Statist. 2004; 53: 665-666. doi: 10. 1111/j. 1467-9876. 2004. 02031. x (Pubitemid 39289328)
38. Ramos D, Franco-Pedroso J, González-Rodríguez J. Calibration and weight of the evidence by human listeners. The ATVS-UAM submission to NIST Human Aided Speaker Recognition 2010. In: Proceedings of the International Conference on Speech Signal Processing, Prague, Czech Republic. IEEE. 2011. p. 5908-5911. doi: 10. 1109/ICASSP. 2011. 5947706
39. Lindh J, Morrison GS. Forensic voice comparison by humans and machine: Forensic voice comparison on a small database of Swedish voice recordings. In: Lee W-S, Zee E, Proceedings of the 17th International Congress of Phonetic Sciences, Hong Kong, China. Hong Kong: Organizers of ICPhS XVII at the Department of Chinese, Translation and Linguistics, City University of Hong Kong. 2011. p. 1254-1257.
40. Morrison GS, Ochoa F, Thiruvaran T. Database selection for forensic voice comparison. In: Proceedings of Odyssey 2012: The Language and Speaker Recognition Workshop, Singapore. International Speech Communication Association. 2012. p. 62-77.