Decision Tree Classifier with default configuration for predictions.
This model predicts the category of the salary of a person based on it's financials information. The categories are '<=50K' or '>50K'
The original dataset is from the UCI Machine Learning Repository. Source: https://archive.ics.uci.edu/ml/datasets/census+income The original data set has 48842 rows, and a 80-20 split was used to break this into a train and test set. No stratification was done. To use the data for training a One Hot Encoder was used on the features and a label binarizer was used on the target class.
20% data is used for evaluation.
Model performance in precision, recall and fbeta are: precision: 0.608 recall: 0.639 fbeta: 0.623
Metics were also calculated on data slices. Saved results are in 'slice_output.txt'. The model may potentially discriminate people so further analysis should be done.
The census.csv used in this project differs from the UCI Machine Learning Repository as it has only 32561 rows.