ML-Classification - Puffins-Classifier

Let's classify some cute puffins

Puffins refer to three species of small alcids in the bird genus Fratercula. For eight months of the year, puffins live on the ocean. These pint-sized birds are well adapted for sea life and enjoy devouring fish. When the warmer spring weather arrives, they head to land for breeding – usually around May to August.

Unfortunately for puffins their good looks fade away at some point each year. One of their most distinct features, the brightly coloured parrot-like beak, loses its technicolour in winter. But when breeding season starts on land, their vibrant appearance returns in time for them to attract a mate.

---

Our data

The data is in puffins.csv. Our data consists of 344 entries, each with six input variables (x) and one output variable (y). Each entry is in one of three classes of puffins (Tufted, Horned or Atlantic). Our goal is to train the classifier to use the inputs and distinguish the species of puffins.

The input variables are:

1. Beak Length (mm)
2. Beak Depth (mm)
3. Wing Length (mm)
4. Body Mass (g)
5. Sex
6. Year of Data Collection

Model Training

Since our training data has labels attached to them, this is a supervised classification problem. In the jupyter notebook, our workflow will be

1. Data Cleaning, Wrangling and Visualization
2. Seperating training and testing set
3. Training and evaluating three classification models
4. Model fine tuning based on training performance
5. Evaluating and comparing testing performance
6. Identify potential areas of improvement

---

References

https://www.rabbies.com/en/blog/everything-you-need-know-about-puffins