On the dependence of predictive models on experimental dataset: a spontaneous combustion studies scenario

ABSTRACT

Spontaneous combustion has been a long-standing challenge affecting coal mines globally as it jeopardises the financial and safety aspect of mining operations. It does not only damage assets but can also injure mineworkers due to inhalation of the noxious emissions or extreme burning. To manage this problem, different experimental methods and predictive models to foresee the propensity of coal to spontaneous combustion were reviewed. The review shows that most predictive models created are dependent on experimental analysis to ascertain their reliability and validity.

KEYWORDS:

• Artificial intelligence
• coalfield
• models
• propensity
• regression
• spontaneous combustion

Disclosure statement

No potential conflict of interest was reported by the author(s).