Predictive Modeling

In this section, we build and evaluate machine learning models to predict the success of a Falcon 9 first stage landing. The objective is to determine the best performing algorithm for future predictions based on the features prepared during data wrangling.

Since the target variable (success) is binary (0 = failure, 1 = success), this is a classification problem. We will test different classification algorithms and compare their performance.

Our modeling process follows these steps:

• Define target variable (success) and feature set (X).
• Split the dataset into training and testing sets (80/20).
• Train multiple classification models:
  • Logistic Regression
  • Support Vector Machines (SVM)
  • Decision Tree Classifier
  • K-Nearest Neighbors (KNN)
  • Random Forest Classifier
• Optimize hyperparameters using GridSearchCV.
• Evaluate models with accuracy scores and confusion matrices.
• Select the best performing model for deployment.