Exploratory analysis of multivariate drill core time series measurements

Demand for mineral resources is increasing, necessitating exploitation of lower grade and more heterogeneous orebodies. The high variability inherent in such orebodies leads to an increase the cost, complexity environmental footprint associated with mining processing. Enhanced knowledge orebody characteristics thus vital companies optimize profitability. We present a pilot study investigate prediction geometallurgical variables from drill sensor data. A comparison made performance multilayer perceptron (MLP) multiple linear regression models (MLR) predicting variable. This based on simulated data that are physically realistic, having been derived fitted one available core. terms mean standard deviation (over repeated samples population) absolute error, root square coefficient determination. best performing model depends form response variable sample size. measures tends be higher MLP, MLR appears offer consistent test cases considered. References R. M. Balabin S. V. Smirnov. Interpolation extrapolation problems multivariate analytical chemistry: Benchmarking robustness near-infrared (NIR) spectroscopy data”. Analyst 137.7 (2012), pp. 1604–1610. doi: 10.1039/c2an15972d C. Bishop. Pattern recognition machine learning. Springer, 2006. url: https://link.springer.com/book/9780387310732 J. B. Boisvert, E. Rossi, K. Ehrig, Deutsch. Geometallurgical modeling at Olympic dam mine, South Australia”. Math. Geosci. 45 (2013), 901–925. 10.1007/s11004-013-9462-5 T. Bollerslev. Generalized autoregressive conditional heteroskedasticity”. Economet. 31.3 (1986), 307–327. 10.1016/0304-4076(86)90063-1 Both Dimitrakopoulos. Applied learning throughput prediction—A case using production Tropicana Gold Mining Complex”. Minerals 11.11 (2021), p. 1257. 10.3390/min11111257 Chen G. Li. Tsallis wavelet entropy its application power signal analysis”. Entropy 16.6 (2014), 3009–3025. 10.3390/e16063009 Coward, Vann, Dunham, Stewart. primary-response framework variables”. Seventh international geology conference. 2009, 109–113. https://www.ausimm.com/publications/conference->url: https://www.ausimm.com/publications/conference- proceedings/seventh-international-mining-geology- conference-2009/the-primary-response-framework-for- geometallurgical-variables/ A. Davis N. Christensen. Derivative analysis layer selection geophysical borehole logs”. Comput. 60 34–40. 10.1016/j.cageo.2013.06.015 Dritsaki. An empirical evaluation GARCH volatility modeling: Evidence Stockholm stock exchange”. Fin. 7.2 (2017), 366–390. 10.4236/jmf.2017.72020 F. Engle Modelling persistence variances”. Econ. Rev. 5.1 1–50. 10.1080/07474938608800095 Hadi Ling. Some cautionary notes use principal components regression”. Am. Statistician 52.4 (1998), 15–19. 10.2307/2685559 Hunt, Kojovic, Berry. Estimating comminution indices ore mineralogy, chemistry core logging”. Second AusIMM International Geometallurgy Conference (GeoMet) 2013. 2013, 173–176. http://ecite.utas.edu.au/89773>url: http://ecite.utas.edu.au/89773 C210). Hyndman, Y. Kang, P. Montero-Manso, Talagala, Wang, Yang, O’Hara-Wild, Ben Taieb, H. Cao, D. Lake, Laptev, Moorman. tsfeatures: Time series feature extraction. R package version 1.0.2. 2020. https://CRAN.R-project.org/package=tsfeatures>url: https://CRAN.R-project.org/package=tsfeatures C222). L. Johnson, Browning, Pendock. Hyperspectral imaging applications geometallurgy: Utilizing blast hole mineralogy predict Au-Cu recovery Phoenix Nevada”. Geol. 114.8 (2019), 1481–1494. 10.5382/econgeo.4684 Martin Morris. overview statistical process control continuous batch monitoring”. Trans. Inst. Meas. Control 18.1 (1996), 51–60. 10.1177/014233129601800107 Sepulveda, Dowd, Xu, Addo. Multivariate modelling by projection pursuit”. 49.1 121–143. 10.1007/s11004-016-9660-z Webb, Cooper, Ashwal. Wavelet investigation density susceptibility Bellevue Moordkopje borehole, Bushveld Complex, Africa”. SEG Technical Program Expanded Abstracts 2008. Society Exploration Geophysicists, 2008, 1167–1171. 10.1190/1.3059129 Zuo. Identifying geochemical anomalies Cu Pb–Zn skarn mineralization component spectrum–area fractal Gangdese Belt, Tibet (China)”. Geochem. Explor. 111.1-2 (2011), 13–22. 10.1016/J.GEXPLO.2011.06.012