加强深度

stock/dnn_train.py

@@ -11,10 +11,10 @@ def read_data(path):
     train_y=[]
     lines = []
     with open(path) as f:
-        for x in range(150000):
+        for x in range(100000):
             lines.append(eval(f.readline().strip()))
     random.shuffle(lines)
-    lines = lines[:20000]
+    lines = lines[:80000]
     d=int(0.95*len(lines))
 
     size = len(lines[0])
@@ -24,18 +24,27 @@ def read_data(path):
     test_y=[s[size-1] for s in lines[d:]]
     return np.array(train_x),np.array(train_y),np.array(test_x),np.array(test_y)
 
-train_x,train_y,test_x,test_y=read_data("D:\\data\\quantization\\stock5.log")
+train_x,train_y,test_x,test_y=read_data("D:\\data\\quantization\\stock6.log")
 
 model = Sequential()
-model.add(Dense(units=325, input_dim=163,  activation='relu'))
+model.add(Dense(units=425, input_dim=166,  activation='relu'))
+model.add(Dense(units=325, activation='relu'))
+model.add(Dense(units=325, activation='relu'))
 model.add(Dense(units=225, activation='relu'))
+model.add(Dense(units=225, activation='relu'))
+model.add(Dense(units=225, activation='relu'))
+model.add(Dense(units=225, activation='relu'))
+model.add(Dense(units=225, activation='relu'))
+model.add(Dense(units=225, activation='relu'))
+model.add(Dense(units=125, activation='relu'))
 model.add(Dense(units=125, activation='relu'))
+model.add(Dense(units=166, activation='relu'))
 # model.add(Dropout(0.2)(Dense(units=225, activation='relu')))
 model.add(Dense(units=8, activation='softmax'))
 model.compile(loss='categorical_crossentropy', optimizer="adam",metrics=['accuracy'])
 
 print("Starting training ")
-h=model.fit(train_x, train_y, batch_size=16, epochs=10, shuffle=True)
+h=model.fit(train_x, train_y, batch_size=32, epochs=8, shuffle=True)
 score = model.evaluate(test_x, test_y)
 print(score)
 print('Test score:', score[0])