Experiential Model

The spider's decision process factors in three elements: the model's certainty, the output value, and the time until the actualization of that output value.

Output as a Probability Distribution

Output as Differentiable with Respect to Inputs

Memoryless-ness

Flexibility as a Black Box

Beyond the above requirements, the experiential model need not be restricted to any particular algorithm or strategy. This can be a neural network, a gaussian mixture model, a bayesian polynomial regression model, etc. etc.