The goal of this project was to create a machine learning solution to recommend the best career based on an individual's personality and aptitude scores. The dataset contains OCEAN personality traits and various aptitude scores for multiple careers.
Upon exploring the dataset, we discovered that most careers have only one sample, making it impossible to train a meaningful ML model.
Instead of ML, we use a similarity-based recommendation system:
- Represent each career as a point in multi-dimensional trait space.
- Compute Euclidean distances from a new individual to all careers.
- Recommend the top N closest careers as the best fit.
Optionally, a KNN classifier with n_neighbors=1 can be used; this is functionally equivalent to the distance-based approach.