Skip to content

Instantly share code, notes, and snippets.

@abinashpanda
Created June 27, 2014 21:59
Show Gist options
  • Save abinashpanda/0181ebcd560b8a006bcd to your computer and use it in GitHub Desktop.
Save abinashpanda/0181ebcd560b8a006bcd to your computer and use it in GitHub Desktop.
Display the source blob
Display the rendered blob
Raw
{
"metadata": {
"name": "",
"signature": "sha256:259ff372e94fe0694951b47f4898bfeddf236842acd72f7d93bc92c56870e3af"
},
"nbformat": 3,
"nbformat_minor": 0,
"worksheets": [
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Comparision of Multilabel Models\n",
"================================\n",
"\n",
"In this notebook, we are going to compare the performance of two multi-label models\n",
"* ```MultilabelModel model``` : with **constant entry term $0$** in joint feature vector to **not** model bias term\n",
"* ```MultilabelModel model_with_bias``` : with **constant entry $1$** in the joint feature vector to model bias term\n",
"\n",
"The joint feature vector are:\n",
"* ```model```$\\leftrightarrow \\psi(x, y) = [x || 0] \\otimes y$\n",
"* ```model_with_bias```$\\leftrightarrow \\psi(x, y) = [x || 1] \\otimes y$\n",
"\n",
"where $\\otimes$ is the $tensor product$.\n",
"\n",
"For comparision of the two models, we are going to perform on the datasets with binary labels. \n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"##Experiment 1 : Binary Label Data\n",
"###Generation of some synthetic data"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"from __future__ import print_function, division\n",
"\n",
"try:\n",
" from sklearn.datasets import make_classification\n",
"except:\n",
" print(\"scikit-learn is required to generate synthetic data sets\")\n",
" \n",
"import numpy as np\n",
"\n",
"X, Y = make_classification(n_samples=1000,\n",
" n_features=2,\n",
" n_informative=2,\n",
" n_redundant=0,\n",
" n_clusters_per_class=2)\n",
"X = X + 1"
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 1
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###Preparation of data and model for Shogun Toolbox"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"from modshogun import RealFeatures\n",
"\n",
"def create_features(X, constant):\n",
" \"\"\"\n",
" Parameters\n",
" ----------\n",
" X : array like\n",
" features of the data set\n",
" constant : int\n",
" constant to be concatenated with\n",
" feature vector\n",
" \n",
" Returns\n",
" -------\n",
" features : RealFeatures\n",
" features compatible with shogun\n",
" \"\"\"\n",
" features = RealFeatures(\n",
" np.c_[X, constant * np.ones(X.shape[0])].T)\n",
" \n",
" return features"
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 2
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"from modshogun import MultilabelSOLabels\n",
"\n",
"def create_labels(Y, n_classes):\n",
" \"\"\"\n",
" Parameters\n",
" ----------\n",
" Y : array or list\n",
" array or list of labels\n",
" n_classes : int\n",
" number of binary label assignment\n",
" per examples\n",
" \n",
" Returns\n",
" -------\n",
" labels : MultilabelSOLabels\n",
" labels compatible with shogun\n",
" \"\"\"\n",
" try:\n",
" n_samples = Y.shape[0]\n",
" except AttributeError:\n",
" n_samples = len(Y)\n",
" \n",
" labels = MultilabelSOLabels(n_samples, n_classes)\n",
" for i, sparse_label in enumerate(Y):\n",
" try:\n",
" sparse_label = sorted(sparse_label)\n",
" except TypeError:\n",
" sparse_label = [sparse_label]\n",
" labels.set_sparse_label(i, np.array(sparse_label, dtype=np.int32))\n",
" \n",
" return labels"
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 15
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"feats_0 = create_features(X, 0)\n",
"feats_1 = create_features(X, 1)\n",
"labels = create_labels(Y, 2)\n",
"\n",
"from modshogun import MultilabelModel\n",
"\n",
"model = MultilabelModel(feats_0, labels)\n",
"model_with_bias = MultilabelModel(feats_1, labels)"
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 16
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###Training and Evaluation of Structured Machines with/without Threshold"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"from modshogun import StochasticSOSVM, DualLibQPBMSOSVM, StructuredAccuracy, LabelsFactory\n",
"from time import time\n",
"\n",
"sgd = StochasticSOSVM(model, labels)\n",
"sgd_with_bias = StochasticSOSVM(model_with_bias, labels)\n",
"\n",
"sgd.train()\n",
"sgd_with_bias.train()"
],
"language": "python",
"metadata": {},
"outputs": [
{
"metadata": {},
"output_type": "pyout",
"prompt_number": 17,
"text": [
"True"
]
}
],
"prompt_number": 17
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"def evaluate_machine(machine,\n",
" X_test,\n",
" Y_test,\n",
" n_classes,\n",
" bias):\n",
" \"\"\"\n",
" Evaluating the accuracy of a structured machine\n",
" \n",
" Parameters\n",
" ----------\n",
" machine : StructuredMachine\n",
" machine to be trained and evaluated\n",
" X_test : array\n",
" array of features\n",
" Y_test : array or list\n",
" array or list of labels to be tested\n",
" n_classes : int\n",
" number of binary class assignment per label\n",
" bias : bool\n",
" if machine is learning the bias or not\n",
" \n",
" Returns\n",
" -------\n",
" accuracy : float\n",
" accuracy\n",
" \"\"\"\n",
" if bias:\n",
" feats_test = create_features(X_test, 1)\n",
" else:\n",
" feats_test = create_features(X_test, 0)\n",
" \n",
" test_labels = create_labels(Y_test, n_classes)\n",
" \n",
" out_labels = LabelsFactory.to_structured(machine.apply(feats_test))\n",
" evaluator = StructuredAccuracy()\n",
" accuracy = evaluator.evaluate(out_labels, test_labels)\n",
" \n",
" return accuracy "
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 18
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"print(\">>> Accuracy without threshold tuning = %f \" % evaluate_machine(sgd, X, Y, 2, False))"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
">>> Accuracy without threshold tuning = 0.827000 \n"
]
}
],
"prompt_number": 48
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"print(\">>> Accuracy with threshold tuning = %f \" %evaluate_machine(sgd_with_bias, X, Y, 2, True))"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
">>> Accuracy with threshold tuning = 0.946000 \n"
]
}
],
"prompt_number": 49
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###Plotting the *Data* along with the *Boundary* "
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"import matplotlib.pyplot as plt\n",
"%matplotlib inline\n",
"\n",
"def plot_hyperplane(machine_0,\n",
" machine_1,\n",
" label_0,\n",
" label_1,\n",
" title,\n",
" X, y):\n",
" \"\"\"\n",
" Parameters\n",
" ----------\n",
" machine_0 : StructuredMachine\n",
" machine for which the separating \n",
" hyperplane is to be plot\n",
" \n",
" machine_1 : StructuredMachine\n",
" another machine for which the separating\n",
" hyperplane is to be plot\n",
" \n",
" label_0 : str\n",
" label for hyper plane of machine_0\n",
" \n",
" label_1 : str\n",
" label for hyper plane of machine_1\n",
"\n",
" title : str\n",
" title of the plot\n",
" \n",
" X : array like\n",
" array of features\n",
" \n",
" y : array or list like\n",
" array of labels\n",
" \"\"\"\n",
" def get_parameters(weights):\n",
" return -weights[0]/weights[1], -weights[2]/weights[1]\n",
" \n",
" zeros_class = np.where([lab == 0 for lab in y])\n",
" ones_class = np.where([lab == 1 for lab in y])\n",
" plt.scatter(X[zeros_class, 0], X[zeros_class, 1], c='b', label=\"Negative Class\")\n",
" plt.scatter(X[ones_class, 0], X[ones_class, 1], c='r', label=\"Positive Class\")\n",
" m_0, c_0 = get_parameters(machine_0.get_w()) \n",
" m_1, c_1 = get_parameters(machine_1.get_w())\n",
" x_min, x_max = np.min(X[:, 0]) - 0.5, np.max(X[:, 0]) + 0.5\n",
" y_min, y_max = np.min(X[:, 1]) - 0.5, np.max(X[:, 1]) + 0.5\n",
" xx = np.linspace(x_min, x_max)\n",
" yy_0 = m_0 * xx + c_0\n",
" yy_1 = m_1 * xx + c_1\n",
" plt.plot(xx, yy_0, \"k--\", label=label_0)\n",
" plt.plot(xx, yy_1, \"g-\", label=label_1)\n",
" plt.legend(loc=\"best\")\n",
" plt.xlim((x_min, x_max))\n",
" plt.ylim((y_min, y_max))\n",
" plt.title(title)\n"
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 50
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"fig = plt.figure(figsize=(10, 10))\n",
"plot_hyperplane(sgd, sgd_with_bias,\n",
" \"Boundary for machine *without* bias for class 0\",\n",
" \"Boundary for machine *with* bias for class 0\",\n",
" \"Binary Classification using SO-SVM with/without threshold tuning\",\n",
" X, Y)"
],
"language": "python",
"metadata": {},
"outputs": [
{
"metadata": {},
"output_type": "display_data",
"png": "iVBORw0KGgoAAAANSUhEUgAAAk0AAAJZCAYAAAC0vQHVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
"text": [
"<matplotlib.figure.Figure at 0x7f104ae90f50>"
]
}
],
"prompt_number": 51
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"##Experiment 2 : Multilabel Data\n",
"###Loading of data from LibSVM File"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"def load_data(file_name):\n",
" \"\"\"\n",
" Parameters\n",
" ----------\n",
" file_name : str\n",
" path of the LibSVM file \n",
" \n",
" Returns\n",
" -------\n",
" X : array like\n",
" features matrix\n",
" Y : list\n",
" list of labels\n",
" n_samples : int\n",
" number of samples\n",
" n_features : int\n",
" number of features\n",
" n_classes : int\n",
" number of class labels\n",
" \"\"\"\n",
" input_file = open(file_name)\n",
" lines = input_file.readlines()\n",
" n_samples = len(lines)\n",
" n_features = len(lines[0].split()) - 1\n",
" Y = []\n",
" X = []\n",
" for line in lines:\n",
" data = line.split()\n",
" Y.append(map(int, data[0].split(\",\")))\n",
" feats = []\n",
" for feat in data[1:]:\n",
" feats.append(float(feat.split(\":\")[1]))\n",
" X.append(feats)\n",
" X = np.array(X)\n",
" n_classes = max(max(label) for label in Y) + 1\n",
" return X, Y, n_samples, n_features, n_classes"
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 53
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###Training and Evaluation of Structured Machines with/without Threshold"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"def test_multilabel_data(train_file,\n",
" test_file):\n",
" \"\"\"\n",
" Parameters\n",
" ----------\n",
" train_file : str\n",
" path to file containing training data\n",
" test_file : str\n",
" path to file containing test data\n",
" \n",
" Returns\n",
" -------\n",
" accuracy : float\n",
" accuracy of multi-label classification\n",
" without threshold tuning\n",
" accuracy_tt : float\n",
" accuracy of multi-label classification\n",
" with threshold tuning\n",
" \"\"\"\n",
" X_train, Y_train, n_samples, n_features, n_classes = load_data(train_file)\n",
"\n",
" X_test, Y_test, n_samples, n_features, n_classes = load_data(test_file)\n",
"\n",
" multilabel_feats_0 = create_features(X_train, 0)\n",
" multilabel_feats_1 = create_features(X_train, 1)\n",
" multilabel_labels = create_labels(Y_train, n_classes)\n",
"\n",
" multilabel_model = MultilabelModel(multilabel_feats_0, multilabel_labels)\n",
" multilabel_model_with_bias = MultilabelModel(multilabel_feats_1, multilabel_labels)\n",
" \n",
" multilabel_sgd = StochasticSOSVM(multilabel_model, multilabel_labels)\n",
" multilabel_sgd_with_bias = StochasticSOSVM(multilabel_model_with_bias, multilabel_labels)\n",
" \n",
" multilabel_sgd.train()\n",
" multilabel_sgd_with_bias.train()\n",
" \n",
" return (evaluate_machine(multilabel_sgd,\n",
" X_test, Y_test,\n",
" n_classes, False),\n",
" evaluate_machine(multilabel_sgd_with_bias,\n",
" X_test, Y_test,\n",
" n_classes, True))\n",
" "
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 68
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"def print_result(train_file,\n",
" test_file):\n",
" print(\n",
" \"\"\"\n",
" >>> Accuracy of multi-label classification *without* threshold tuning = %f\n",
" >>> Accuracy of multi-label classification *with* threshold tuning = %f\n",
" \"\"\" % (\n",
" test_multilabel_data(train_file, test_file)))"
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 75
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###[Yeast Multi-Label Data](http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/multilabel.html)"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"print_result(\"/home/abinash/shogun/data/multilabel/yeast_train.svm\",\n",
" \"/home/abinash/shogun/data/multilabel/yeast_test.svm\")"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"\n",
" >>> Accuracy of multi-label classification *without* threshold tuning = 0.339940\n",
" >>> Accuracy of multi-label classification *with* threshold tuning = 0.491876\n",
" \n"
]
}
],
"prompt_number": 76
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###[Scene Multi-Label Data](http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/multilabel.html)"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"print_result(\"/home/abinash/shogun/data/multilabel/scene_train\",\n",
" \"/home/abinash/shogun/data/multilabel/scene_test\")"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"\n",
" >>> Accuracy of multi-label classification *without* threshold tuning = 0.548774\n",
" >>> Accuracy of multi-label classification *with* threshold tuning = 0.579571\n",
" \n"
]
}
],
"prompt_number": 77
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###Comparision with scikit-learn's implementation"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"from sklearn.multiclass import OneVsRestClassifier\n",
"from sklearn.svm import SVC\n",
"from sklearn.metrics import jaccard_similarity_score\n",
"from sklearn.preprocessing import MultiLabelBinarizer\n",
"\n",
"def sklearn_implementation(train_file,\n",
" test_file):\n",
" label_binarizer = MultiLabelBinarizer()\n",
"\n",
" X_train, Y_train, n_samples, n_features, n_classes = load_data(train_file)\n",
" X_test, Y_test, n_samples, n_features, n_classes = load_data(test_file)\n",
"\n",
" clf = OneVsRestClassifier(SVC(kernel='linear'))\n",
" clf.fit(X_train, label_binarizer.fit_transform(Y_train))\n",
" print(\">>> Accuracy of multilabel classification using scikit-learn = %f\" % (\n",
" jaccard_similarity_score(label_binarizer.fit_transform(Y_test), clf.predict(X_test))))"
],
"language": "python",
"metadata": {},
"outputs": [],
"prompt_number": 79
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###[Yeast Multi-Label Data](http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/multilabel.html)"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"sklearn_implementation(\"/home/abinash/shogun/data/multilabel/yeast_train.svm\",\n",
" \"/home/abinash/shogun/data/multilabel/yeast_test.svm\")"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
">>> Accuracy of multilabel classification using scikit-learn = 0.497962\n"
]
}
],
"prompt_number": 80
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"###[Scene Multi-Label Data](http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/multilabel.html)"
]
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"sklearn_implementation(\"/home/abinash/shogun/data/multilabel/scene_train\",\n",
" \"/home/abinash/shogun/data/multilabel/scene_test\")"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
">>> Accuracy of multilabel classification using scikit-learn = 0.576226\n"
]
}
],
"prompt_number": 81
}
],
"metadata": {}
}
]
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment