Merge pull request #323 from Neuroblox/fix_striatal_learning

striatal learning rule modified

src/blox/reinforcement_learning.jl

@@ -34,6 +34,8 @@ end
 mutable struct HebbianModulationPlasticity <: AbstractLearningRule
     const K
     const decay
+    const α
+    const θₘ
     state_pre
     state_post
     t_pre
@@ -42,11 +44,11 @@ mutable struct HebbianModulationPlasticity <: AbstractLearningRule
     modulator
 
     function HebbianModulationPlasticity(; 
-        K, decay, modulator=nothing,
+        K, decay, α, θₘ, modulator=nothing,
         state_pre=nothing, state_post=nothing, 
         t_pre=nothing, t_post=nothing, t_mod=nothing,   
     )
-        new(K, decay, state_pre, state_post, t_pre, t_post, t_mod, modulator)
+        new(K, decay, α, θₘ, state_pre, state_post, t_pre, t_post, t_mod, modulator)
     end
 end
 
@@ -55,8 +57,10 @@ dlogistic(x) = logistic(x) * (1 - logistic(x))
 function (hmp::HebbianModulationPlasticity)(val_pre, val_post, val_modulator, w, feedback)
     DA = hmp.modulator(val_modulator, feedback)
     DA_baseline = hmp.modulator.κ_DA * hmp.modulator.N_time_blocks
-
-    Δw = hmp.K * val_post * val_pre * DA * (DA - DA_baseline) * dlogistic(DA) - hmp.decay * w
+    ϵ = feedback - (hmp.modulator.κ_DA - DA)
+
+   # Δw = hmp.K * val_post * val_pre * DA * (DA - DA_baseline) * dlogistic(DA) - hmp.decay * w
+    Δw = hmp.K * val_post * val_pre * ϵ * (ϵ + hmp.θₘ) * dlogistic(hmp.α * (ϵ + hmp.θₘ)) - hmp.decay * w
 
     return Δw
 end

test/reinforcement_learning.jl

@@ -28,7 +28,7 @@ using CSV
     bloxs = [VAC, PFC, STR_L, STR_R, SNcb, TAN_pop, AS, stim]
     d = Dict(b => i for (i,b) in enumerate(bloxs))
 
-    hebbian_mod = HebbianModulationPlasticity(K=0.2, decay=0.01, modulator=SNcb, t_pre=t_trial, t_post=t_trial, t_mod=0.31*t_trial)
+    hebbian_mod = HebbianModulationPlasticity(K=0.2, decay=0.01, α=3, θₘ=1, modulator=SNcb, t_pre=t_trial, t_post=t_trial, t_mod=0.31*t_trial)
     hebbian = HebbianPlasticity(K=0.2, W_lim=2, t_pre=t_trial, t_post=t_trial)
 
     g = MetaDiGraph()