Reference · The Applied Higher-Order Functions catalogue

rainbow.algorithms.proofs.ahof.mc-dropout

The Applied Higher-Order Functions cataloguesrc/rainbow/algorithms/proofs/ahof/mc_dropout.clj1 definitions, 4 theorems

AHoF V62 MC Dropout Uncertainty: estimate predictive uncertainty from the variance of stochastic-dropout forward passes and flag an input uncertain when that variance reaches the threshold: uncertain? variance threshold = decide (threshold <= variance). Proved native and axiom-free (decision, two positive instances, at-most certificate, negative instance). Requires audit/decide-nnrat-at-most, nnrat/at-most, nnrat/decide-at-most-true-sound (all in the D63 checkpoint). Scope (measured by the kernel): this module has 4 worked instances (statements about specific values). None of its theorems is a guarantee about the algorithm for all inputs; `lein verify rainbow.algorithms.proofs.ahof.mc-dropout` prints this classification.

mc-dropout/uncertain? :
  (x₁ x₂ : ℚ≥0)
  → data/Bool

mc-dropout/high-variance-uncertain :
  → bo/is-safe 3/4 1/2 = data/true

mc-dropout/boundary-uncertain :
  → bo/is-safe 1/2 1/2 = data/true

mc-dropout/variance-certificate :
  → 1/2 ≤ 3/4

mc-dropout/low-variance-confident :
  → bo/is-safe 1/4 1/2 = data/false