Recruitment Decision Analysis Dataset

Data Description

The dataset includes the following columns:

Variable Description

• Age:

  • Range: 20 to 50 years
  • Type: Integer

• Gender:

  • Categories: Male (0) or Female (1)
  • Type: Binary

• Education Level:

  • Categories: 1: Bachelor (type 1), 2: Bachelor (type 2), 3: Master, 4: Doctorate
  • Type: Categorical

• Years of Work Experience:

  • Range: 0 to 15 years
  • Type: Integer

• Number of Previously Worked Companies:

  • Range: 1 to 5 companies
  • Type: Integer

• Distance to Company:

  • Range: 1 to 50 km
  • Type: Continuous Float

• Interview Score:

  • Range: 0 to 100
  • Type: Integer

• Skill Score:

  • Range: 0 to 100
  • Type: Integer

• Personality Score:

  • Range: 0 to 100
  • Type: Integer

• Recruitment Strategy:

  • Categories: 1: Aggressive, 2: Moderate, 3: Conservative
  • Type: Categorical

• Recruitment Decision (Target Variable):

  • Categories: 0: Not hired, 1: Hired
  • Type: Binary Integer

Exploratory Data Analysis (EDA)

Correlation Analysis

Correlation analysis identifies relationships between features and the target variable (recruitment decision), revealing the most influential factors in the hiring process.

Age and Recruitment Decision

Analysis of age distribution and its impact on recruitment decisions uncovers age‑related trends and biases, using histograms and box plots for different age groups.

Years of Work Experience and Recruitment Decision

Exploration of how years of work experience affect the likelihood of being hired, visualized with scatter and line plots.

Predictive Modeling

Logistic Regression

A logistic regression model is employed to predict recruitment decisions. Logistic regression is suitable for this binary classification problem, where the target outcome is either hired or not hired.

Hyperparameter Tuning

GridSearchCV is used for hyperparameter optimization, tuning parameters such as regularization penalty (penalty: l1, l2) and regularization strength (C: [0.01, 0.1, 1, 10, 100]).