The iris is a widely cultivated perennial plant with 310 accepted species. Due to its showy flowers, the it is often grown as an ornamental plant, which makes their cultivation of commercial interest. Customers, it would be imagined, would like as much information about the plant they are buying as possible. But the species of iris can be rather difficult to ascertain, and is often mislabeled by growers to increase profit yields. Therefore, we explore here methods by which we can use machine learning to classify incoming flowers based on measurements taken from the plant. With any luck, the process can be automated given a well-performing model.

For the task, we use Fisher's iris dataset as training data for our models. This dataset contains the length and width measurements of the sepal and petal of three species of iris (Iris setosa, Iris virginica, Iris versicolor). We will first explore the data, after which techniques will be evaluated by which we can classify the species of iris, given its measurements.