add simple nn

ale3otik · Dec 2, 2017 · d511045 · d511045
1 parent 9417936
commit d511045
Show file tree

Hide file tree

Showing 2 changed files with 44 additions and 10 deletions.
diff --git a/models/dnn.py b/models/dnn.py
@@ -1,5 +1,5 @@
 import keras.backend as K
-from keras.layers import Dense, Activation, Input, LSTM, Dropout, multiply, Lambda
+from keras.layers import Dense, Activation, Input, LSTM, Dropout, multiply, Lambda, BatchNormalization
 from keras.layers import RNN
 from keras.models import Model, Sequential
 from keras import optimizers
@@ -13,27 +13,30 @@
 
 def nn_1(input_shape=None, drop_val=0.1, eps_reg=1e-2, hidden_units=200):
     model = Sequential()
-    model.add(Dense(hidden_units=hidden_units, input_shape=(input_shape,),
+    model.add(Dense(units=hidden_units, 
+                    input_shape=(input_shape,),
                     kernel_initializer='normal',
-                    kernel_regularizer=regularizers.l2(0.01)
-                    activation=tanh,
+                    kernel_regularizer=regularizers.l2(eps_reg),
+                    activation=tanh
                    ))
 
     model.add(Dropout(drop_val))
     # model.add(LeakyReLU(alpha=0.01))
     model.add(Dense(units=hidden_units,
                     kernel_initializer='normal',
-                    kernel_regularizer=regularizers.l2(eps_reg)
+                    kernel_regularizer=regularizers.l2(eps_reg),
+                    activation=tanh
                    ))
     model.add(Dropout(0.1))
     model.add(BatchNormalization())
 
     model.add(Dense(units=1,
                     kernel_initializer='normal',
                     kernel_regularizer=regularizers.l2(eps_reg),
-                    activation=linear,
+                    activation=linear
                    ))
-    optimizer = optimizers.Adam(optimizers.Adam(1e-3, clipvalue=10.0))
+
+    optimizer = optimizers.Adam(1e-3, clipvalue=10.0)
     model.compile(loss='mean_squared_error', optimizer=optimizer)
     return model
 

diff --git a/models/keras_lstm_usage.ipynb b/models/keras_lstm_usage.ipynb
@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 2,
    "metadata": {
     "collapsed": false
    },
@@ -12,6 +12,37 @@
     "import keras.layers"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "(100, 10)\n",
+      "(100, 1)\n",
+      "Epoch 1/20"
+     ]
+    }
+   ],
+   "source": [
+    "import dnn\n",
+    "import numpy as np\n",
+    "import pandas as pd \n",
+    "reload(dnn)\n",
+    "# model = keras_lstm.keras_lstm_model_1(max_sequence_length=100,input_shape=10)\n",
+    "model = dnn.nn_1(input_shape=10)\n",
+    "x = np.random.random(1000).reshape(100,10)\n",
+    "y = np.random.random(100).reshape(100,1)\n",
+    "print(x.shape)\n",
+    "print(y.shape)\n",
+    "model.fit(x=x,y=y, epochs=20, verbose=1,batch_size=5)"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 4,
@@ -174,7 +205,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "Python [default]",
+   "display_name": "Python 3",
    "language": "python",
    "name": "python3"
   },
@@ -188,7 +219,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.3"
+   "version": "3.5.3"
   }
  },
  "nbformat": 4,