CIS$470$ Intro to Data Analytics

Lab Title: Logistic Regression in Python: Building Your First Model

Estimated Time: $2$ hours

Learning Objectives

By the end of this lab, you will be able to:

Load and preprocess data for logistic regression.

Understand and apply logistic regression for binary classification.

Evaluate the model's performance using key metrics like accuracy, precision, recall, and ROC$-$AUC.

Prerequisites

This lab assumes familiarity with Python programming and basic data analysis using libraries such as pandas and numpy. No prior knowledge of regression models is required.

$$

You may need to install packages such as sklearn $($Sci kit learn using commands such as the below. I ran this in my vscode terminal on a windows PC$.$ Please use VS Code.

Dataset

We$$ll use the Titanic dataset, a popular dataset for binary classification tasks. The Titanic dataset contains information about passengers, such as age, sex, class, and survival status. Our goal is to predict the likelihood of a passenger surviving based on these features.

Dataset Info:

Target Variable: Survived $(1$ if the passenger survived, $0$ if not$)$

Features: Pclass, Sex, Age, SibSp, Parch, Fare, Embarked

$$

$$

$$

$$

$$

Lab Steps

Step $1$: Import Required Libraries

In this step, you will import the essential libraries needed for data manipulation, visualization, and model building.

Step $2$: Load and Explore the Dataset

Load the Titanic dataset.

Check the dataset$$s structure and identify missing values or data types that need processing.

Step $3$: Data Preprocessing

Handle Missing Values: Fill in missing values for the Age column with the median age.

Convert Categorical Variables: Convert the Sex and Embarked columns into dummy variables.

Select Features and Target: Choose relevant columns for prediction.

Step $4$: Split the Data

Split the dataset into training and testing sets.

Step $5$: Build the Logistic Regression Model

Initialize the logistic regression model.

Fit the model to the training data.

$$

Step $6$: Make Predictions

Use the trained model to make predictions on the test data.

Step $7$: Evaluate the Model

Calculate evaluation metrics to assess model performance:

Accuracy: Proportion of correct predictions.

Precision: How often the model is correct when it predicts survival.

Recall: How often the model correctly identifies survivors.

ROC$-$AUC: The area under the ROC curve.

$$

Step $8$: Visualize Results

Confusion Matrix: Plot the confusion matrix to understand the model$$s performance in terms of true$/$false positives and negatives.

ROC Curve: Plot the ROC curve to visualize the model's performance.

Step $9$:

$$

$$

$$

$$

Interpretation and Conclusion

Discuss the following:

$1.$ The significance of each metric $($accuracy$,$ precision, recall, ROC$-$AUC$)$ and how it reflects the model$$s performance.

$2.$ How well the logistic regression model performed and potential limitations of using a logistic regression model

$3.$ Possible steps for improving the model, such as adding more features, tuning hyperparameters, or testing different classification methods.