Predictive Modeling on Real Hiring Data: Exploring Salary Stereotypes and Bias Mitigation Documentation

Welcome to the documentation for Predictive Modeling on Real Hiring Data: Exploring Salary Stereotypes and Bias Mitigation.

Main Notebook:

• Load the data
• Remove rows
  • Drop duplicates
  • Clean the columns’ names
  • Create new IDs and separate different people with duplicating IDs
  • Extract numer_of_searches column
  • Remove irrelevant columns
  • Remove candidates in first stages
  • Removal of Candidates with Inconsistent Final Outcomes
• Preprocess columns
  • Ral Mapping
  • Overall mapping
  • Make the Protected category column boolean
  • Remove invalid values from Job Title Hiring
  • Aggregate Records
  • Residence
  • Save Preprocessed Dataframe
• Create dataset
  • Load Preprocessed Dataframe
  • Custom Similarity Features
  • Dataset Analysis
  • Drop Nan Values
• Training
  • Load Cleaned Dataset
  • Models Comparison
  • Features Removal Comparison
  • Sampling Techniques
  • Fairness Metrics
  • Removing Protected Atttributes

Inprocessing Bias Mitigation Techniques Notebook:

• Inprocessing Bias Mitigation Techniques
  • Load Data
  • Base Model
  • Fauci
  • Prejudice Remover

Preprocessing Bias Mitigation Techniques Notebook:

• Preprocessing Bias Mitigation Techniques
  • Load Data
  • Model
  • Disparate Impact Remover
  • Reweighing

Utilities:

• Feature Engineering Utilities
  • calculate_distance()
  • calculate_experience_match_score()
  • calculate_professional_similarity_score()
  • calculate_salary_fit_score()
  • calculate_study_area_score()
  • calculate_study_title_score()
  • create_candidate_text()
  • create_job_text()
  • prepare_nlp_text_columns()
• Data Cleaning Utilities
  • clean_dataframe_columns()
  • remove_initial_stage_candidates()
  • remove_not_hired_valid_candidates()
  • sort_group()
  • split_duplicate_ids_by_invariant_columns()
• Plotting Utilities
  • get_mean_std()
  • plot_metrics()
  • plot_metrics_grouped()
  • print_fairness_results_table()

Indices and tables

• Index

• Module Index

• Search Page