FurkanKovan / morethan101

morethan101 Data Analysis Bootcamp w/Ezgi Turalı, provided by DeveloperMultiGroup

Data Analysis & LGBM Model w/Ezgi Turalı (morethan101)

To review the file paste the link to nbviewer.jupyter.org.

This notebook is provided by Ezgi Turalı, all credit to her.

• twitter.com/ezgiturali

Thank you to DeveloperMultiGroup for helding #morethan101 Data Analysis bootcamp.

• linkedin.com/company/developermultigroup/
• twitter.com/devmultigroup
• instagram.com/devmultigroup

Briefly, About

In this notebook, we analyzed key features of the 'spaceship-titanic' dataset from the Kaggle competition. First the features are visualized and analyzed. Later by comparing various machine learning models, with the help of LazyPredict library, we saw that LGBM and RandomForest models are the most convenient. We picked the LGBM model to make our submission.

Results

• LGBM Accuracy: %81.08
• RandomForest Accuracy: %74.0
• Kaggle Competition Score: 0.80453