Hands-On Automated Machine Learning

Visualizing high-dimensional datasets

What about visually inspecting datasets that have more than three dimensions? In order to visually inspect your dataset, you need to have a maximum of three dimensions; if not, you need to use specific methods to reduce dimensionality. This is usually achieved by applying a Principal Component Analysis (PCA) or t-SNE algorithm.

The following code will load the Breast Cancer Wisconsin Diagnostic dataset, which is commonly used in ML tutorials:

# Wisconsin Breast Cancer Diagnostic Datasetfrom sklearn.datasets import load_breast_cancerimport pandas as pddata = load_breast_cancer()X = data.datadf = pd.DataFrame(data.data, columns=data.feature_names)df.head()

Output in the console is as follows:

 mean radius ...