Poshmark - Take-Home assignment

This repo consists of notebook with details on data exploration, modelling, evaluation, and a script to generate model-agnostic feature importances

This link contains a presentation with overview of my approach and answering some questions that they asked.

Files in archive

• attr{1,2,3,4,5}.json: The categorical mappings for the features in the data set
• feat_imp.py: code to generate instance level feature importances
• poshmark_notebook.ipynb: Noteboook with all details of modeling feature engineering. exploration etc.
• req.txt: Environment File with package information for creating model environment
• feat_imps_1.json: Instance level feature importances calculated with the model built.
• model_pipeline.joblib: Model pipeline with all preprocessing steps and the model
• poshmark_notebook.ipynb: The presentation with answers and explanation for the questions asked in pdf