hw1: wip

assignment_1/deliverable/run_model.py

@@ -1,6 +1,10 @@
+#!/usr/bin/env python3
+
 import joblib
 import numpy as np
-
+from keras import models
+import sys
+from keras.models import Model
 
 def load_data(filename):
     """
@@ -57,13 +61,51 @@ if __name__ == '__main__':
     # EDITABLE SECTION OF THE SCRIPT: if you need to edit the script, do it here
     ############################################################################
 
-    # Load the trained model
-    baseline_model_path = './baseline_model.pickle'
-    baseline_model = load_model(baseline_model_path)
+    ms = {"baseline", "linear", "nonlinear"}
 
-    # Predict on the given samples
-    y_pred = baseline_model.predict(x)
+    if len(sys.argv) != 2:
+        print("%s {model_name}" % sys.argv[0])
+        sys.exit(1)
 
+    if sys.argv[1] not in ms:
+        print("model name must be one of " + str(ms))
+        sys.exit(2)
+
+    y_pred = None
+    if sys.argv[1] == "baseline":
+        # Load the trained model
+        baseline_model_path = './baseline_model.pickle'
+        baseline_model = load_model(baseline_model_path)
+
+        # Predict on the given samples
+        y_pred = baseline_model.predict(x)
+    else:
+        X = np.zeros([x.shape[0], 5])
+        X[:, 0] = 1
+        X[:, 1:3] = x
+        X[:, 3] = x[:, 0] * x[:, 1]
+        X[:, 4] = np.sin(x[:, 0])
+
+        if sys.argv[1] == "linear":
+            # Load the linear model
+            linear_model_path = './linear_regression.pickle'
+            linear_model = load_model(linear_model_path)
+
+            # Predict on the given samples
+            y_pred = linear_model.predict(X)
+        else:
+            # Load saved Keras model
+            nonlinear_model = models.load_model('./nonlinear_model')
+            # Load saved normalizing factors for Xs
+            normalizers = load_model('./nonlinear_model_normalizers.pickle')
+            # Normalize Xs with normalizing factors
+            X = X[:, 1:3]
+            X -= normalizers["mean"]
+            X /= normalizers["std"]
+
+            # Run model on normalized data
+            y_pred = nonlinear_model.predict(X)
+            y_pred = y_pred[:, 0]
 
     ############################################################################
     # STOP EDITABLE SECTION: do not modify anything below this point.

assignment_1/src/build_models.py

@@ -2,14 +2,18 @@ import numpy as np
 from sklearn.linear_model import LinearRegression
 from sklearn.model_selection import train_test_split
 from tensorflow.keras import Input, Model
-from tensorflow.keras.models import Sequential
+from tensorflow.keras.models import Sequential, save_model
 from tensorflow.keras.layers import Dense
 from tensorflow.keras import losses
 from tensorflow import keras
 import tensorflow as tf
 from sklearn.metrics import mean_squared_error
+from utils import save_sklearn_model, save_keras_model
+import os
 
-data = np.load("../data/data.npz")
+d = os.path.dirname(__file__)
+
+data = np.load(d + "/../data/data.npz")
 xs = data["x"] # 2000x2
 y = data["y"] # 2000x1
 points = 2000
@@ -43,6 +47,8 @@ print("# Linear regression:")
 # Print the resulting parameters
 print("f(x) = %g + %g * x_1 + %g * x_2 + %g * x_1 * x_2 + %g * sin(x_1)" % tuple(reg.coef_))
 
+save_sklearn_model(reg, d + "/../deliverable/linear_regression.pickle")
+
 # Test using MSQ on test set
 score = reg.score(X_test, y_test)
 print("MSQ error on test set is: %g" % (score))
@@ -53,48 +59,45 @@ print("\n# Feed-forward NN:")
 
 A = X_val
 
-X_train = X_train[:, 1:]
-X_val = X_val[:, 1:]
+X_train = X_train[:, 1:3]
+X_val = X_val[:, 1:3]
 
 # X_train = X_train[:, 1:3]
 # X_val = X_val[:, 1:3]
 
 mean = np.mean(X_train, axis=0)
 std = np.std(X_train, axis=0)
-#y_mean = np.mean(y_train, axis=0)
-#y_std = np.std(y_train, axis=0)
-
-y_mean = 0
-y_std = 1
 
 X_train -= mean
 X_train /= std
-y_train -= y_mean
-y_train /= y_std
 
 X_val -= mean
 X_val /= std
-y_val -= y_mean
-y_val /= y_std
 
 network = Sequential()
-network.add(Dense(35, activation='relu'))
+network.add(Dense(30, activation='relu'))
+network.add(Dense(20, activation='relu'))
+network.add(Dense(20, activation='relu'))
 network.add(Dense(10, activation='relu'))
+network.add(Dense(5, activation='relu'))
 network.add(Dense(3, activation='relu'))
+network.add(Dense(2, activation='relu'))
 network.add(Dense(1, activation='linear'))
 network.compile(optimizer='rmsprop', loss='mse', metrics=['mse'])
 
-epochs = 10000
-callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=1000)
-network.fit(X_train, y_train, epochs=epochs, verbose=1, batch_size=100, validation_data=(X_val, y_val), callbacks=[callback])
+epochs = 100000
+callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=40)
+network.fit(X_train, y_train, epochs=epochs, verbose=1, batch_size=15,
+        validation_data=(X_val, y_val), callbacks=[callback])
+
+network.save(d + "/../deliverable/nonlinear_model")
+save_sklearn_model({"mean": mean, "std": std}, d + "/../deliverable/nonlinear_model_normalizers.pickle")
 
 msq = mean_squared_error(network.predict(X_val), y_val)
 print(msq)
 
-X_test = X_test[:, 1:]
+X_test = X_test[:, 1:3]
 X_test -= mean
 X_test /= std
-y_test -= y_mean
-y_test /= y_std
 msq = mean_squared_error(network.predict(X_test), y_test)
 print(msq)