From a8fe22801cb7abdca543ea6b592a7b058475d5de Mon Sep 17 00:00:00 2001
From: Pawel Morawiecki <pawel.morawiecki@gmail.com>
Date: Wed, 10 May 2017 11:20:39 +0200
Subject: [PATCH] Short tutorial how to evaluate test examples
---
for_investigation.ipynb | 188 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 188 insertions(+), 0 deletions(-)
create mode 100644 for_investigation.ipynb
diff --git a/for_investigation.ipynb b/for_investigation.ipynb
new file mode 100644
index 0000000..3e3e95e
--- /dev/null
+++ b/for_investigation.ipynb
@@ -0,0 +1,188 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "Using TensorFlow backend.\n"
+ ]
+ }
+ ],
+ "source": [
+ "from keras.models import Model\n",
+ "from keras.layers import Input, Dense, Dropout, Activation, BatchNormalization\n",
+ "from keras.optimizers import SGD, Adam\n",
+ "import numpy as np"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ "filename = 'test_set.csv'\n",
+ "raw_data = open(filename, 'rt')\n",
+ "test_data = np.loadtxt(raw_data, delimiter= '\\t')"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ "number_of_features = 1126\n",
+ "test_set = test_data[:,0:1126]\n",
+ "test_labels = test_data[:,1126] #last column consists of labels"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Neural network configuration"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ "inputs = Input(shape=(number_of_features,))\n",
+ "output_from_1st_layer = Dense(1000, activation='relu')(inputs)\n",
+ "output_from_1st_layer = Dropout(0.5)(output_from_1st_layer)\n",
+ "output_from_1st_layer = BatchNormalization()(output_from_1st_layer)\n",
+ "output_from_2nd_layer = Dense(500, activation='relu')(output_from_1st_layer)\n",
+ "output_from_2nd_layer = Dropout(0.5)(output_from_2nd_layer)\n",
+ "output_from_2nd_layer = BatchNormalization()(output_from_2nd_layer)\n",
+ "output = Dense(1, activation='sigmoid')(output_from_2nd_layer)\n",
+ "\n",
+ "model = Model(inputs, output)\n",
+ "model.compile(optimizer='Adam',loss='binary_crossentropy',metrics=['accuracy'])"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Let's load weights learnt earlier"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ "model.load_weights(\"weights_2017_05_10.h5\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Evaluation"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "First, calculate predictions for test set"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ " predictions = model.predict(test_set)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Now we can calculate basic metrics"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Accuracy:0.7316259444607988\n",
+ "Precision: 0.7378337531486147\n",
+ "Recall: 0.7185752134236091\n",
+ "F1: 0.7280771525154107\n"
+ ]
+ }
+ ],
+ "source": [
+ " true_positives = 0.0\n",
+ " false_positives = 0.0\n",
+ " true_negatives = 0.0\n",
+ " false_negatives = 0.0\n",
+ "\n",
+ " for i in range(len(test_set)):\n",
+ " if (predictions[i]<0.5 and test_labels[i]==0): true_negatives += 1 \n",
+ " if (predictions[i]<0.5 and test_labels[i]==1): false_negatives += 1\n",
+ " if (predictions[i]>=0.5 and test_labels[i]==1): true_positives += 1\n",
+ " if (predictions[i]>=0.5 and test_labels[i]==0): false_positives += 1 \n",
+ " \n",
+ " accuracy = (true_positives+true_negatives)/len(test_set)\n",
+ " precision = true_positives/(true_positives+false_positives)\n",
+ " recall = true_positives/(true_positives+false_negatives)\n",
+ " f1 = 2*(precision*recall)/(precision+recall)\n",
+ "\n",
+ " print ('Accuracy:' + repr(accuracy))\n",
+ " print ('Precision: ' + repr(precision))\n",
+ " print ('Recall: ' + repr(recall))\n",
+ " print ('F1: ' + repr(f1))"
+ ]
+ }
+ ],
+ "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": "2.7.6"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
--
libgit2 0.22.2