Mathematical Geosciences

, Volume 47, Issue 5, pp 521–545 | Cite as

Classification of Gold-Bearing Particles Using Visual Cues and Cost-Sensitive Machine Learning

  • Tom Horrocks
  • Daniel Wedge
  • Eun-Jung Holden
  • Peter Kovesi
  • Nick Clarke
  • John Vann
Article
First Online:
Received:
Accepted:

Abstract

Ore sorting increases the grade of an ore feed stream by separating very low-grade particles (‘waste’) from those containing higher concentrations of the desired mineral (‘ore’), thus economically reducing the amount of material processed in further mineral concentration steps. This paper reports a preliminary study that aims to develop an automated method for discriminating waste and gold-bearing particles. The study used both hyperspectral measurements and RGB images of waste and gold-bearing particles from the Sunrise Dam Gold Mine as input to the discriminating method. Advanced feature extraction methods were employed to capture visual cues such as texture and colour from the RGB images, which were combined with hyperspectral features to give nine types of representative features. Feature selection was applied to groups of the representative features and resulting feature subsets were evaluated using three machine learning algorithms, namely a support vector machine, a naïve Bayes classifier, and a majority decision table, to identify a highly informative subset of features. Cost-sensitive training was used to minimise the nominal profit lost due to sorting error based on real cost values from the milling process, with the aim of economically balancing the ore acceptance rate with the waste rejection rate. A cost-blind support vector machine achieved an ore acceptance rate of 84 % and a waste rejection rate of 87 %, which resulted in $0.98 nominal profit lost per tonne of crushed rock particles. Cost-sensitive training reduced the nominal profit lost to $0.34 per tonne, undercutting the costs associated with refining all particles by $0.24 per tonne.

Keywords

Gold Ore classification Support vector machine  MetaCost Hyperspectral reflectance Feature selection 
This is a preview of subscription content, log in to check access.

Notes

Acknowledgments

We would like to thank the funding body AngloGold Ashanti Australia, and note that a co-author, John Vann, was VP Mineral Resources for this company at the time the work was done.

References

  1. ASD Inc (2008) TerraSpec user manual: 600541 rev. gGoogle Scholar
  2. Berman M, Bischof L, Huntington J (1999) Algorithms and software for the automated identification of minerals using field spectra or hyperspectral imagery. In: 13th international conference on applied geologic remote sensing, pp 222–232Google Scholar
  3. Bradley AP (1997) The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognit 30:1145–1159CrossRefGoogle Scholar
  4. Chatterjee S, Bandopadhyay S, Machuca D (2010) Ore grade prediction using a genetic algorithm and clustering based ensemble neural network model. Math Geosci 42(3):309–326. doi: 10.1007/s11004-010-9264-y CrossRefGoogle Scholar
  5. Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20(3):273–297. doi: 10.1023/A:1022627411411 Google Scholar
  6. Domingos P (1999) MetaCost: a general method for making classifiers cost-sensitive. In: Proceedings of the 5th ACM SIGKDD international conference on knowledge discovery and data mining, ACM, New York, KDD ’99, pp 155–164. doi: 10.1145/312129.312220
  7. Domingos P, Pazzani M (1997) On the optimality of the simple Bayesian classifier under zero-one loss. Mach Learn 29(2–3):103–130CrossRefGoogle Scholar
  8. Fawcett T (2006) An introduction to ROC analysis. Pattern Recognit Lett 27(8):861–874. doi: 10.1016/j.patrec.2005.10.010 CrossRefGoogle Scholar
  9. Gama J, de Carvalho A (2009) Machine learning. IGI Global, book section, Hershey, pp 2462–2468. doi: 10.4018/978-1-60566-026-4.ch392
  10. Hall MA (1999) Correlation-based feature selection for machine learning. PhD thesis, The University of WaikatoGoogle Scholar
  11. Hall M, Holmes G (2003) Benchmarking attribute selection techniques for discrete class data mining. IEEE Trans Knowl Data Eng 15(6):1437–1447. doi: 10.1109/TKDE.2003.1245283 CrossRefGoogle Scholar
  12. Hall M, Frank E, Holmes G, Pfahringer B, Reutemann P, Witten IH (2009) The WEKA data mining software: an update. ACM SIGKDD Explor Newsl 11(1):10–18CrossRefGoogle Scholar
  13. Holden EJ, Fu SC, Kovesi P, Dentith M, Bourne B, Hope M (2011) Automatic identification of responses from porphyry intrusive systems within magnetic data using image analysis. J Appl Geophys 74(4):255–262. doi: 10.1016/j.jappgeo.2011.06.016 CrossRefGoogle Scholar
  14. Holden EJ, Wong JC, Kovesi P, Wedge D, Dentith M, Bagas L (2012) Identifying structural complexity in aeromagnetic data: an image analysis approach to greenfields gold exploration. Ore Geol Rev 46:47–59. doi: 10.1016/j.oregeorev.2011.11.002 CrossRefGoogle Scholar
  15. Huang J, Ling CX (2005) Using AUC and accuracy in evaluating learning algorithms. IEEE Trans Knowl Data Eng 17:299–310CrossRefGoogle Scholar
  16. Hunt EB, Marin J, Stone PJ (1966) Experiments in induction. Academic Press, New YorkGoogle Scholar
  17. John GH, Kohavi R, Pfleger K (1994) Irrelevant features and the subset selection problem. In: Proceedings of the 11th international conference on machine learning, Morgan Kaufmann, Burlington, pp 121–129Google Scholar
  18. Kistner M, Jemwa GT, Aldrich C (2013) Monitoring of mineral processing systems by using textural image analysis. Miner Eng 52:169–177. doi: 10.1016/j.mineng.2013.05.022 CrossRefGoogle Scholar
  19. Kohavi R (1995) The power of decision tables. In: Lavrac N, Wrobel S (eds) Machine learning: ECML-95. Lecture notes in computer science, vol 912. Springer, Berlin, pp 174–189. doi: 10.1007/3-540-59286-5_57
  20. Kononenko I (1994) Estimating attributes: analysis and extensions of RELIEF. In: Proceedings of the European conference on machine learning on machine learning, ECML-94, pp 171–182. Springer, New YorkGoogle Scholar
  21. Lewis DD (1998) Naive (Bayes) at forty: the independence assumption in information retrieval. In: Machine learning: ECML-98, pp 4–15. Springer, New YorkGoogle Scholar
  22. Liu H, Motoda H (2007) Computational methods of feature selection. Chapman & Hall/CRC Data Mining and Knowledge Discovery Series, Taylor & Francis, New YorkGoogle Scholar
  23. Liu H, Setiono R (1996) A probabilistic approach to feature selection—a filter solution. In: ICML96, pp 319–327Google Scholar
  24. Loy G, Zelinsky A (2003) Fast radial symmetry for detecting points of interest. IEEE Trans Pattern Anal Mach Intell 25(8):959–973CrossRefGoogle Scholar
  25. Maron ME (1961) Automatic indexing: an experimental inquiry. J ACM 8(3):404–417. doi: 10.1145/321075.321084 CrossRefGoogle Scholar
  26. Marsden JO, House CI (2006) Chemistry of gold extraction. SME, EnglewoodGoogle Scholar
  27. Matheron G (1984) The selectivity of the distributions and “the second principle of geostatistics”. In: Verly G, David M, Journel A, Marechal A (eds) Geostatistics for natural resources characterization, pp 421–433. Springer, The Netherlands. doi: 10.1007/978-94-009-3699-7_24
  28. Nugus M, Briggs M, Tombs S, Elms P, Erickson M (2013) Sunrise dam gold mine, AngloGold Ashanti. In: Australasian mining and metallurgical operating practices: the Sir Maurice Mawby memorial volume. Monograph, vol 2, 3rd edn. The Australian Institute of Mining and Metallurgy, Carlton, p 1920Google Scholar
  29. Ojala T, Pietikinen M, Harwood D (1996) A comparative study of texture measures with classification based on featured distributions. Pattern Recognit 29(1):51–59. doi: 10.1016/0031-3203(95)00067-4 CrossRefGoogle Scholar
  30. Perez CA, Estvez PA, Vera PA, Castillo LE, Aravena CM, Schulz DA, Medina LE (2011) Ore grade estimation by feature selection and voting using boundary detection in digital image analysis. Int J Miner Process 101(14):28–36. doi: 10.1016/j.minpro.2011.07.008 CrossRefGoogle Scholar
  31. Platt JC (1999) Fast training of support vector machines using sequential minimal optimization. In: Advances in kernel methods. MIT Press, Cambridge, pp 185–208Google Scholar
  32. Rendu JM (2008) Introduction to cut-off grade estimation. Society for Mining, Metallurgy, and Exploration (SME), EnglewoodGoogle Scholar
  33. Singh V, Rao SM (2005) Application of image processing and radial basis neural network techniques for ore sorting and ore classification. Miner Eng 18(15):1412–1420. doi: 10.1016/j.mineng.2005.03.003 CrossRefGoogle Scholar
  34. Tahmasebi P, Hezarkhani A (2012) A hybrid neural networks-fuzzy logic–genetic algorithm for grade estimation. Comput Geosci 42:18–27. doi: 10.1016/j.cageo.2012.02.004 CrossRefGoogle Scholar
  35. Tessier J, Duchesne C, Gauthier C, Dufour G (2008) Estimation of alumina content of anode cover materials using multivariate image analysis techniques. Chem Eng Sci 63(5):1370–1380. doi: 10.1016/j.ces.2007.07.028 CrossRefGoogle Scholar

Copyright information

© International Association for Mathematical Geosciences 2015

Authors and Affiliations

  • Tom Horrocks
    • 1
  • Daniel Wedge
    • 1
  • Eun-Jung Holden
    • 1
  • Peter Kovesi
    • 1
  • Nick Clarke
    • 2
  • John Vann
    • 1
    • 3
  1. 1.Centre for Exploration Targeting, School of Earth and EnvironmentThe University of Western AustraliaCrawleyAustralia
  2. 2.AngloGold Ashanti Australia LimitedPerthAustralia
  3. 3.Anglo American PLCLondonUnited Kingdom

Personalised recommendations