dnn_TensorFlow.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Brain Age Prediction using Deep Neural Network with TensorFlow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda/lib/python2.7/site-packages/statsmodels/compat/pandas.py:56: FutureWarning: The pandas.core.datetools module is deprecated and will be removed in a future version. Please use the pandas.tseries module instead.\n",
      "  from pandas.core import datetools\n",
      "/anaconda/lib/python2.7/site-packages/h5py/__init__.py:34: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import random as rn\n",
    "import pandas as pd\n",
    "\n",
    "from plotnine import * \n",
    "\n",
    "# for reproducibility results:\n",
    "np.random.seed(42)\n",
    "\n",
    "import tensorflow as tf\n",
    "from keras import backend as K\n",
    "tf.set_random_seed(42)\n",
    "sess = tf.Session(graph=tf.get_default_graph())\n",
    "K.set_session(sess)\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation, Dropout\n",
    "# from keras.layers.core import Dense, Activation\n",
    "from keras.optimizers import SGD\n",
    "from keras import optimizers\n",
    "from keras import regularizers\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "#from keras.callbacks import TensorBoard\n",
    "import time\n",
    "\n",
    "from sklearn import preprocessing\n",
    "from keras.wrappers.scikit_learn import KerasRegressor\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import mean_absolute_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data\n",
    "dat = np.load('./full_reduced/full/feaDAT.npy')\n",
    "age = np.load('./full_reduced/full/ageDAT.npy')\n",
    "sex = np.load('./full_reduced/full/sexDAT.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# preprocess data\n",
    "X = preprocessing.scale(dat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2026, 5124)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Splitting into Training set and Test set, X, Y pair\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, age, test_size = 0.2, random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "#model setup\n",
    "N = 50;\n",
    "actfun = 'tanh' # 'relu'\n",
    "def create_model():\n",
    "    model = Sequential()\n",
    "    model.add(Dense(N, input_dim=X.shape[1], kernel_initializer='normal', \n",
    "                    kernel_regularizer=regularizers.l2(0.01), \n",
    "                    activation = 'tanh'))\n",
    "    model.add(Dropout(0.50))\n",
    "    model.add(Dense(N, kernel_initializer='normal', \n",
    "                    kernel_regularizer=regularizers.l2(0.01), \n",
    "                    activation= 'tanh'))\n",
    "    model.add(Dropout(0.50))\n",
    "    model.add(Dense(N, kernel_initializer='normal', \n",
    "                    kernel_regularizer=regularizers.l2(0.01), \n",
    "                    activation= 'tanh'))\n",
    "    model.add(Dropout(0.50))\n",
    "    model.add(Dense(N, kernel_initializer='normal', \n",
    "                    kernel_regularizer=regularizers.l2(0.01), \n",
    "                    activation= 'tanh')) \n",
    "    model.add(Dropout(0.50))\n",
    "    model.add(Dense(N, kernel_initializer='normal', \n",
    "                    kernel_regularizer=regularizers.l2(0.01), \n",
    "                    activation= 'tanh'))\n",
    "    \n",
    "    model.add(Dropout(0.50))\n",
    "    model.add(Dense(1, kernel_initializer='normal'))\n",
    "    model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "    return model\n",
    "              \n",
    "# evaluate model with standardized dataset\n",
    "estimator = KerasRegressor(build_fn = create_model, epochs=1000, batch_size=100, verbose=0, shuffle=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DNN model fitted in 200.792 s\n"
     ]
    }
   ],
   "source": [
    "t0 = time.time()\n",
    "baseline_history = estimator.fit(X_train, Y_train)\n",
    "dnn_fit = time.time() - t0\n",
    "print(\"DNN model fitted in %.3f s\" % dnn_fit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train data correlation 0.985 \n"
     ]
    }
   ],
   "source": [
    "y_train_pred = estimator.predict(X_train)\n",
    "corr_train = np.corrcoef(Y_train, y_train_pred)[0,1]\n",
    "print(\"Train data correlation %.3f \" % corr_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test data correlation 0.927 \n"
     ]
    }
   ],
   "source": [
    "X_prediction = estimator.predict(X_test)\n",
    "corr_test = np.corrcoef(Y_test, X_prediction)[0,1]\n",
    "print(\"Test data correlation %.3f \" % corr_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.377745162024286"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_absolute_error(Y_test, X_prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEWCAYAAABxMXBSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(baseline_history.history['loss'])\n",
    "#plt.plot(baseline_history.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[min, median, max] age: [6 25 88] years\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAh0AAAFzCAYAAACadrtQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<ggplot: (7582322553)>\n"
     ]
    }
   ],
   "source": [
    "# bias shown in brain age gap: over/under-estimate for lower/upper ends\n",
    "idx = np.argsort(Y_test)\n",
    "age_gap = X_prediction - Y_test\n",
    "dframe = pd.DataFrame()\n",
    "dframe['age_gap'] = age_gap\n",
    "dframe['Y_test'] = Y_test\n",
    "p2 = ggplot(dframe, aes('Y_test','age_gap')) + geom_point() +\\\n",
    "    geom_hline(yintercept=[0], size=1, linetype='--') +\\\n",
    "    xlab('chronological age') +\\\n",
    "    ylab('brain age gap') +\\\n",
    "    stat_smooth(method='lm') \n",
    "  \n",
    "print(\"[min, median, max] age: [%d %d %d] years\" % (min(Y_test), np.median(Y_test), max(Y_test)))\n",
    "print(p2) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('dnn.pickle', 'wb') as f:\n",
    "    pickle.dump([Y_test, age_gap], f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}