1. Data Loading and Column Selection

• 1.1 Load the data
• 1.2 Clean Column Names
• 1.3 Select relevant columns

2. Data Cleaning and Formatting

• 2.1 Identify missing values
• 2.2 Drop missing values
• 2.3 Data type conversions
• 2.4 Convert qualified_candidates and candidate_hired to numeric
• 2.5 Convert difficulty_level to numeric
• 2.6 Double checking the data types

3. Feature Engineering

• 3.1 Budget allocation percentages
• 3.2 Cost per application (CPA) per channel
• 3.3 Why CPA was just used for visualization later on

4. Exploratory Data Patterns

• 4.1 Cross-validation with StratifiedKFold
• 4.1 PCA 2D Visualization
• 4.2 K-Means Clustering
• 4.3 Cluster vs. Outcome Analysis