Closing in on something reasonable for roads

1 files changed, 48 insertions, 36 deletions

diff --git a/src/train5.py b/src/train5.py
index 3718fd0..ea25e5e 100644
--- a/src/train5.py
+++ b/src/train5.py
@@ -1,53 +1,65 @@
-import tensorflow as tf
 import pandas as pd
 import numpy as np
 
+from tensorflow.keras.regularizers import l1
+from tensorflow.keras.regularizers import l2
+
+import tensorflow.keras.losses
+
 from tensorflow.keras.models import Sequential
-from tensorflow.keras.layers import BatchNormalization
-from tensorflow.keras.layers import Conv2D
-from tensorflow.keras.layers import MaxPooling2D
-from tensorflow.keras.layers import Activation
-from tensorflow.keras.layers import Flatten
 from tensorflow.keras.layers import Dropout
+from tensorflow.keras.layers import Input
 from tensorflow.keras.layers import Dense
+from tensorflow.keras.layers import Convolution2D
+from tensorflow.keras.layers import MaxPooling2D
+from tensorflow.keras.layers import Flatten
+from tensorflow.keras.layers import Activation
+from tensorflow.keras.layers import BatchNormalization
 
-model = Sequential([
-    # Conv2D(5, kernel_size=5, padding='same', input_shape=(5, 5, 2)),
-    # MaxPooling2D(pool_size=2, strides=2),
-    # Activation("relu"),
-    # Flatten(),
-    # Dense(25, activation="relu"),
-    # Dense(10, activation="relu"),
-    # Dense(1, activation="sigmoid")
+# 30 10(11) 5
+# 27 17 4
+
+shape = (-1, 5, 5, 8)
 
-    Dense(30, activation="relu", input_shape=(52,)),  # 30
-    Dense(10, activation="relu"),                     # 10
-    Dense(8, activation="relu"),                      # 8
-    Dense(1, activation="sigmoid")
+model = Sequential([
+    Input(shape[1:]),
+    Convolution2D(1, kernel_size=3, strides=(1, 1),
+                  padding='same',
+                  activation="relu", use_bias=True),
+  # Dropout(0.2),
+    Flatten(),
+    Dense(10, activation="relu", use_bias=True),
+  # Dropout(0.2),
+  # Dense(10, activation="relu", use_bias=False),
+  # Dropout(0.2),
+    Dense(2, activation="softmax")
 ])
 
-model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
+        # bias_regularizer=l2(0.001),
+        # kernel_regularizer=l1(0.001)),
 
-model.summary()
+model.compile(loss='categorical_crossentropy',
+              optimizer='adam',
+              metrics=['accuracy'])
 
-csv = pd.read_csv("data/train5.csv")
+model.summary()
 
-training_data = csv.copy().head(40000)
-training_outcome = training_data.pop('Outcome')
+csv = pd.read_csv("training-5.csv")
 
-training_input = training_data.copy()
-# training_input = training_input.drop(columns=training_data.keys()[0:2])
-train = np.array(training_input)
-# train = train.reshape((training_input.shape[0], 5, 5, 2))
+training_data = csv.head(100000).copy()
+training_input = np.array(training_data.iloc[:, :-2]).reshape(shape)
+training_outcome = training_data.iloc[:, -2:]
 
-model.fit(train, training_outcome, epochs=100)
+test_data = csv.tail(20000).copy()
+test_input = np.array(test_data.iloc[:, :-2]).reshape(shape)
+test_outcome = test_data.iloc[:, -2:]
 
-test_data = csv.copy().tail(50000)
-# test_data = test_data.drop(columns=test_data.keys()[0:2])
-test_outcome = test_data.pop('Outcome')
-test_input = np.array(test_data)
-# test_input = test_input.reshape((test_data.shape[0], 5, 5, 2))
-test_loss, test_acc = model.evaluate(test_input,  test_outcome, verbose=2)
-print('\nTest accuracy:', test_acc)
+results = []
+for i in range(0, 1):
+    print("Iteration ", i)
+    history = model.fit(training_input, training_outcome, epochs=10,
+                        validation_data=(test_input, test_outcome))
+    loss, accuracy = model.evaluate(test_input, test_outcome)
+    results += [(i, loss, accuracy)]
 
-# Test accuracy: 0.67448
+print("\n\nResults: ", results)