There were forest fires before humans existed. Does that mean all forest fires are natural?
The way you answer the question is to look for forest fires that were caused by people. One way would be to use induction to imagine a scenario where huamns start fires, which would leave distinctive evidence that wouldn't happen in natural fires, then look for that, and follow-up. (e.g. fire starts near a campground, evidence of runaway campfire, progression of fire is away from campground, guy who camped there admits "okay, okay, yes, we fell asleep and when we woke up the trees were on fire, and we got scared and ran").
Okay, so, one forest fire, is that a big deal?
The way you answer that question is to become more familiar with forests and forest fires. Maybe a lot of them start near campgrounds? If you investigate and study and learn, maybe you get a sense of how many are caused by people, what the historical record shows about fire rates and scales in the past, etc. Computers might not tell you exactly which trees are going to burn in a new fire you've just discovered, but maybe you can refine the model enough that it can provide a good sense of where it's going to go, how quickly it might expand, where are sparks going to cause new fires, which of these 10 spots should we drop crews, etc.