We briefly introduced the concept of partitioning a dataset into separate datasets for training and testing in Chapter 1, Giving Computers the Ability to Learn from Data, and Chapter 3, A Tour of Machine Learning Classifiers Using Scikit-learn. Remember that the test set can be understood as the ultimate test of our model before we let it loose on the real world. In this section, we will prepare a new dataset, the Wine dataset. After we have preprocessed the dataset, we will explore different techniques for feature selection to reduce the dimensionality of a dataset.
The Wine dataset is another open-source dataset that is available from the UCI machine learning repository (https://archive.ics.uci.edu/ml/datasets/Wine); it consists of 178 wine samples with 13 features describing their different chemical properties.
Using the pandas library, we will directly read in the open source Wine dataset from the UCI machine learning repository:
>>> df_wine = pd.read_csv('https://archive.ics.uci.edu/ml/machine-learning-databases/wine/wine.data', header=None)
>>> df_wine.columns = ['Class label', 'Alcohol',
... 'Malic acid', 'Ash',
... 'Alcalinity of ash', 'Magnesium',
... 'Total phenols', 'Flavanoids',
... 'Nonflavanoid phenols',
... 'Proanthocyanins',
... 'Color intensity', 'Hue',
... 'OD280/OD315 of diluted wines',
... 'Proline']
>>> print('Class labels', np.unique(df_wine['Class label']))
Class labels [1 2 3]
>>> df_wine.head()The 13 different features in the Wine dataset, describing the chemical properties of the 178 wine samples, are listed in the following table:
The samples belong to one of three different classes, 1, 2, and 3, which refer to the three different types of grapes that have been grown in different regions in Italy.
A convenient way to randomly partition this dataset into a separate test and training dataset is to use the train_test_split function from scikit-learn's cross_validation submodule:
>>> from sklearn.cross_validation import train_test_split >>> X, y = df_wine.iloc[:, 1:].values, df_wine.iloc[:, 0].values >>> X_train, X_test, y_train, y_test = ... train_test_split(X, y, test_size=0.3, random_state=0)
First, we assigned the NumPy array representation of feature columns 1-13 to the variable X, and we assigned the class labels from the first column to the variable y. Then, we used the train_test_split function to randomly split X and y into separate training and test datasets. By setting test_size=0.3 we assigned 30 percent of the wine samples to X_test and y_test, and the remaining 70 percent of the samples were assigned to X_train and y_train, respectively.
Note
If we are dividing a dataset into training and test datasets, we have to keep in mind that we are withholding valuable information that the learning algorithm could benefit from. Thus, we don't want to allocate too much information to the test set. However, the smaller the test set, the more inaccurate the estimation of the generalization error. Dividing a dataset into training and test sets is all about balancing this trade-off. In practice, the most commonly used splits are 60:40, 70:30, or 80:20, depending on the size of the initial dataset. However, for large datasets, 90:10 or 99:1 splits into training and test subsets are also common and appropriate. Instead of discarding the allocated test data after model training and evaluation, it is a good idea to retrain a classifier on the entire dataset for optimal performance.