🚄 Shinkansen Travel Experience Hackathon Highlights 🚀
✨ Hackathon Journey: Participated in the #hackofalltrades Hackathon, a thrilling experience packed with invaluable learning.
🎯 Problem Statement: Tasked with predicting passenger delight during Shinkansen (Bullet Train) travel based on a comprehensive survey dataset from 95K passengers, focusing on binary classification for "Overall Experience."
📊 EDA Insights:
Balanced Dependent Variable.
Addressed extensive missing values through strategic imputation.
Merged Survey and Travel datasets using passenger 'ID.'
Minimal outliers judiciously handled.
🔍 Data Preprocessing:
Imputation: Leveraged highest frequency for categorical values.
Advanced imputation based on correlated parameters.
Applied label and one-hot encoding, followed by z-score scaling.
🛠 Modeling Approaches:
Explored various classification algorithms: CART, Random Forest, Boosting, Bagging, Naive Bayes, and Logistic Regression.
Rigorous parameter tuning, cross-validation, and threshold adjustments.
🏆 Top-Performing Model:
Adaptive Boosting with Random Forest as the base estimator emerged as the most effective model.
Achieved an outstanding accuracy of 95.17%.
📈 Leaderboard Ranking: Secured the 9th position on the leaderboard.
✅ Outcome: Successfully crafted a robust predictive model for passenger delight, showcasing expertise in data preprocessing, modeling, and accuracy optimization.
Excited to delve deeper into the world of data science and share more insights! #DataScience #HackathonSuccess #MachineLearning #PredictiveModeling #ShinkansenExperience 😊
The Best Dressed Man In Tech Sales| Helping AE’s Stand Out From 99% of Sellers | On a Quest to Retire Momma 🙏
1moDJ 🐱🐱🐱🐱🐱 🔥🔥🔥 this is so dope!!!