Explore a comprehensive walkthrough of a Kaggle competition starter notebook in this 26-minute video tutorial by Kaggle Grandmaster Rob Mulla. Learn how to approach the PogChamps community challenge, focusing on predictive modeling and Python coding. Discover techniques for feature creation, cross-validation, and building a LightGBM model. Follow along as Rob guides you through loading the starter notebook, analyzing training and test datasets, defining cross-validation schemes, and performing feature engineering. Gain insights into model training, evaluating out-of-fold scores, and interpreting feature importance. Conclude by creating a submission and checking the competition leaderboard. Access additional resources, including the competition link, live coding streams, and related tutorials on pandas and exploratory data analysis.
Overview
Syllabus
Introduction
Competition Page
Loading the Starter Notebook
Training Dataset Description
Test Dataset
Distribution of Target Variable
Defining Cross Validation Scheme
Feature Engineering
Test and Out of Fold Dataframe
Model Training Loop
Training the Model
Evaluating out of fold score
Feature Importance
Create Submission
Checking the Leaderboard and Conclusion
Taught by
Rob Mulla
Related Courses
-
Winning a Kaggle Competition in Python
-
Getting started with TensorFlow 2
-
How to Solve Santander Kaggle Transaction Competition - Top 1% Solution, No Ensemble
-
Understand data science for machine learning
-
Machine Learning - Testing and Error Metrics
-
Instrumenting Weights & Biases for PII Data Detection - ML Pipeline Tutorial
