🧪 Logistic Regression Model Evaluation – Airbnb Listings

📌 Overview

In this lab, I worked on the evaluation phase of the machine learning life cycle, using Logistic Regression to solve a classification problem with the Airbnb "listings" dataset. I defined the problem, preprocessed the data, trained models, performed hyperparameter tuning, evaluated the models with various metrics, and saved my final model for future use.

---

🗂️ What I Did

1. Built My DataFrame & Defined the ML Problem

   • Loaded the Airbnb dataset.
   • Chose a label (target variable) to predict.
   • Identified and preprocessed the features I needed.
   • Created labeled examples and split the data into training and test sets.

2. Trained & Evaluated a Baseline Model

   • Trained a Logistic Regression model using default settings from scikit-learn.
   • Evaluated the model using metrics like accuracy, precision, recall, F1-score, and the confusion matrix.

3. Tuned Hyperparameters with Grid Search

   • Used GridSearchCV to find the best value for C, the regularization strength.
   • Trained a new Logistic Regression model using the optimal C value.
   • Compared this model's performance against the baseline.

4. Plotted Evaluation Curves

   • Plotted the Precision-Recall Curve and the ROC Curve for both models.
   • Calculated and compared the AUC (Area Under the Curve) values.

5. Selected Important Features

   • Used techniques like Recursive Feature Elimination (RFE) and SelectKBest to choose the most important features.
   • Retrained and tested the model using only these features to see how performance changed.

6. Saved the Final Model

   • Made the model persistent using joblib so I can reuse it later for inference without retraining.

---

🛠️ Tools and Libraries I Used

• Python 3.x
• pandas
• numpy
• matplotlib
• seaborn
• scikit-learn
• joblib / pickle

---

🧪 How to Run This Lab

1. Install the necessary libraries:

   pip install pandas numpy matplotlib seaborn scikit-learn

2. Open the notebook:

   jupyter notebook lab_logistic_regression.ipynb

3. Go through each step and run the cells to see the model in action.

---

📁 Files Included

• lab_logistic_regression.ipynb – My complete lab notebook
• listings.csv – The Airbnb dataset
• logistic_model.pkl – The saved logistic regression model
• README.md – This file

---

✅ What I Learned

By completing this lab, I learned how to:

• Frame a classification problem and prepare real-world data
• Train and evaluate logistic regression models
• Tune hyperparameters using grid search
• Visualize performance using precision-recall and ROC curves
• Perform feature selection to improve models
• Save models for future use