In sklearn, all machine learning models are implemented as Python classes from sklearn.linear_model import LogisticRegression train_img = pca.transform(train_img) test_img = pca.transform(test_img) Apply Logistic Regression to the Transformed Data Apply the mapping (transform) to both the training set and the test set. In this case, 95% of the variance amounts to 330 principal components. Note: You can find out how many components PCA choose after fitting the model using pca.n_components_. Note: you are fitting PCA on the training set only. from composition import PCA # Make an instance of the Model pca = PCA(.95)įit PCA on training set. It means that scikit-learn choose the minimum number of principal components such that 95% of the variance is retained. 95 for the number of components parameter. train_img = ansform(train_img) test_img = ansform(test_img) Import and Apply PCA scaler.fit(train_img) # Apply transform to both the training set and the test set. from sklearn.preprocessing import StandardScaler scaler = StandardScaler() # Fit on training set only. If you want to see the negative effect not scaling your data can have, scikit-learn has a section on the effects of not standardizing your data. Note you fit on the training set and transform on the training and test set. StandardScaler helps standardize the dataset’s features. You can transform the data onto unit scale (mean = 0 and variance = 1) which is a requirement for the optimal performance of many machine learning algorithms. PCA is effected by scale so you need to scale the features in the data before applying PCA. The text in this paragraph is almost an exact copy of what was written earlier. from sklearn.model_selection import train_test_split # test_size: what proportion of original data is used for test set train_img, test_img, train_lbl, test_lbl = train_test_split( mnist.data, mnist.target, test_size=1/7.0, random_state=0) Standardize the Data The code below performs a train test split which puts 6/7th of the data into a training set and 1/7 of the data into a test set. The features are 784 dimensional (28 x 28 images) and the labels are simply numbers from 0–9. The labels (the integers 0–9) are contained in mnist.target. The images that you downloaded are contained in mnist.data and has a shape of (70000, 784) meaning there are 70,000 images with 784 dimensions (784 features). from sklearn.datasets import fetch_openml mnist = fetch_openml('mnist_784') You can also add a data_home parameter to fetch_mldata to change where you download the data.
The MNIST database of handwritten digits is more suitable as it has 784 feature columns (784 dimensions), a training set of 60,000 examples, and a test set of 10,000 examples. For this section, we aren’t using the IRIS dataset as the dataset only has 150 rows and only 4 feature columns. While there are other ways to speed up machine learning algorithms, one less commonly known way is to use PCA. pca.explained_variance_ratio_ PCA to Speed-up Machine Learning Algorithms Together, the two components contain 95.80% of the information. By using the attribute explained_variance_ratio_, you can see that the first principal component contains 72.77% of the variance and the second principal component contains 23.03% of the variance. This is important as while you can convert 4 dimensional space to 2 dimensional space, you lose some of the variance (information) when you do this. The explained variance tells you how much information (variance) can be attributed to each of the principal components. With that, let’s get started! If you get lost, I recommend opening the video below in a separate tab. The second part uses PCA to speed up a machine learning algorithm (logistic regression) on the MNIST dataset. To understand the value of using PCA for data visualization, the first part of this tutorial post goes over a basic visualization of the IRIS dataset after applying PCA. Another common application of PCA is for data visualization. This is probably the most common application of PCA. If your learning algorithm is too slow because the input dimension is too high, then using PCA to speed it up can be a reasonable choice. A more common way of speeding up a machine learning algorithm is by using Principal Component Analysis (PCA). One of the things learned was that you can speed up the fitting of a machine learning algorithm by changing the optimization algorithm. My last tutorial went over Logistic Regression using Python. Original image (left) with Different Amounts of Variance Retained
0 kommentar(er)
