From 1533156cba3cebcd115dede8d412efc2e2c90a62 Mon Sep 17 00:00:00 2001
From: S4nd3sh <34370505+S4nd3sh@users.noreply.github.com>
Date: Sun, 11 Mar 2018 10:58:29 +0530
Subject: [PATCH] FeedForward_NeuralNetwork

Down Counter using FeedForwardNeuralNetwork
---
 Neural Networks/DownCounter.R        |  37 +++++++++
 Neural Networks/FF_Neural_Networks.R | 111 +++++++++++++++++++++++++++
 2 files changed, 148 insertions(+)
 create mode 100644 Neural Networks/DownCounter.R
 create mode 100644 Neural Networks/FF_Neural_Networks.R

diff --git a/Neural Networks/DownCounter.R b/Neural Networks/DownCounter.R
new file mode 100644
index 0000000..8e7452d
--- /dev/null
+++ b/Neural Networks/DownCounter.R	
@@ -0,0 +1,37 @@
+
+source("FF_Neural_Networks.R")
+
+#Set layers for Neural Net
+layers <- c(4,9,4)
+
+#Counter Input
+cntr_inp <- as.matrix(rev(expand.grid(c(0,1),c(0,1),c(0,1),c(0,1))))
+
+#Counter Output
+cntr_out <- cntr_inp[nrow(cntr_inp):1,]
+
+
+
+#Invoke Neural Net function
+set.seed(243)
+Output <- Neural.Net(layers = layers,inp = cntr_inp,out = cntr_out,epoch = 100,
+                     step.size = 0.8,error.threshold = 0.001)
+
+#Predict for one state
+cntr_inp[2,]
+round(Neural.Net.Predict(Output,matrix( cntr_inp[2,] ,nrow = 1))[[1]])
+
+#DownCount Prediction 
+while(TRUE){
+  for(i in 1:nrow(cntr_inp)){
+    
+    cat("Down Counter: ")
+    cat("  \b\b")
+    cat(strtoi(paste(round(unlist(Neural.Net.Predict(Output,matrix(cntr_inp[i,],nrow = 1)))),collapse = ""),base = 2))
+    Sys.sleep(0.8)
+    cat("\r")
+    
+    
+  }
+}
+
diff --git a/Neural Networks/FF_Neural_Networks.R b/Neural Networks/FF_Neural_Networks.R
new file mode 100644
index 0000000..e2a645c
--- /dev/null
+++ b/Neural Networks/FF_Neural_Networks.R	
@@ -0,0 +1,111 @@
+#Sigmoid Function
+sigmoid <- function(z){
+  1.0/(1.0+exp(-z))  
+}
+
+#Derivative of Sigmoid
+sigmoid.derivative <- function(z){
+  z*(1-z)
+}
+
+#Randomly Generate Weights
+create.weights <- function(layers){
+  weights <- vector("list",length(layers)-1)
+  for(i in 1:length(weights)){
+    weights[[i]] <- matrix(runif(layers[i]*layers[i+1]),nrow = layers[i],ncol = layers[i+1])
+  }
+  return(weights)
+}
+
+#Randomly Generate Biases
+create.biases <- function(layers){
+  biases <- vector("list",length(layers)-1)
+  for(i in 1:length(biases)){
+    biases[[i]] <- runif(layers[i+1])
+  }
+  return(biases)
+}
+
+#Feedforward / Forward Propogation
+feedforward <- function(inp,weights,biases){
+  layer <- vector("list",length(layers)-1)
+  for(i in 1:(length(layers)-1)){
+    if(i==1){
+      layer[[i]] <- matrix(apply(inp,1,function(x){x %*% weights[[i]] + biases[[i]]}),nrow = nrow(inp),byrow = T)
+      layer_dim  <- dim(layer[[i]])
+      layer[[i]] <- matrix(sigmoid(as.numeric(layer[[i]])),nrow = layer_dim[1],ncol = layer_dim[2],byrow = FALSE)
+    }else{
+      layer[[i]] <- matrix(apply(layer[[i-1]],1,function(x){x %*% weights[[i]] + biases[[i]]}),nrow = nrow(layer[[i-1]]),byrow = T)
+      layer_dim  <- dim(layer[[i]])
+      layer[[i]] <- matrix(sigmoid(as.numeric(layer[[i]])),nrow = layer_dim[1],ncol = layer_dim[2],byrow = FALSE)
+    }
+  }
+  return(layer)
+}
+
+#Calculating Delta Values
+calculate_delta <- function(net_out,out,weights){
+  Slope <- lapply(net_out,sigmoid.derivative)
+  delta <- vector("list",length(Slope))
+  for(i in length(Slope):1){
+    if(i == length(Slope)){
+      delta[[i]] <-  (out-net_out[[i]]) * Slope[[i]]
+    }else{
+      Error_hidden <- delta[[i+1]] %*% t(weights[[i+1]])
+      delta[[i]] <- Error_hidden * Slope[[i]]
+    }
+  }
+  return(delta)
+}
+
+#Updating Weights
+updating_weights <- function(weights,inp,delta,net_out,step.size){
+  for(i in length(weights):1){
+    if(i != 1){
+      weights[[i]] <- weights[[i]] + t(net_out[[i-1]]) %*% delta[[i]] * step.size
+    }else{
+      weights[[i]] <- weights[[i]] + t(inp) %*% delta[[i]] * step.size
+    }
+    
+  }
+  return(weights)
+}
+
+#Updating Biases
+updating_biases <- function(biases,delta,step.size){
+  for(i in 1:length(biases)){
+    biases[[i]] <- biases[[i]] + colSums(delta[[i]]) * step.size
+  }
+  return(biases)
+}
+
+
+#Complete Neural Net operations (Forward and Backward Propogation)
+Neural.Net <- function(layers = NULL,inp,out,epoch,step.size = 0.01,error.threshold = 0.01){
+  weights <- create.weights(layers)
+  biases <- create.biases(layers)
+  for(i in 1:epoch){
+    
+    net_out <- feedforward(inp,weights,biases)
+    
+    ##Back Propogation
+    delta <- calculate_delta(net_out,out,weights)
+    weights <- updating_weights(weights,inp,delta,net_out,step.size)
+    biases <- updating_biases(biases,delta,step.size)
+    avg_error <- mean(abs(out - net_out[[length(net_out)]]))
+    
+    cat(paste("Epoch :",i," | Avg_error = ",avg_error,"\n"))
+    cat("\b\b\r")
+
+    if(avg_error <= error.threshold){
+      break
+    }
+  }
+  return(list(weights = weights,biases = biases,Net_out = net_out))
+}
+
+#Neural Net Prediction
+Neural.Net.Predict <- function(Model,Test.inp){
+  layer <- feedforward(inp = Test.inp,weights = Model$weights,biases = Model$biases)
+  return(layer[length(layer)])
+}