Predicting Corporate Bankruptcy in Poland

This project focuses on building predictive models for corporate bankruptcy using financial data from Polish companies. The objective is to predict corporate bankruptcy.

The notebook covers the following key aspects:

1. Data Loading and Exploration: Initial steps involve loading the dataset and performing exploratory data analysis (EDA) to understand its structure, distributions, and potential issues.

2. Preprocessing: Handles missing values, outliers, and data scaling.

3. Feature Engineering: May involve creating new features from existing ones to improve model performance.

4. Model Selection: Explores various machine learning models suitable for classification tasks (e.g., Logistic Regression, Decision Trees, Random Forests, Gradient Boosting, SVMs).

5. Training and Evaluation: Trains the selected models on the preprocessed data and evaluates their performance using appropriate metrics (e.g., accuracy, precision, recall, F1-score, ROC-AUC), often considering the imbalanced nature of bankruptcy datasets.

6. Hyperparameter Tuning: Optimizes model parameters to achieve better performance.

7. Insights: Aims to provide insights into the factors contributing to corporate bankruptcy and the effectiveness of different predictive models.

In essence, it’s a comprehensive machine learning project demonstrating the process of building and evaluating bankruptcy prediction models using real-world financial data.