Project Details

The challenge at hand involves developing a robust pricing model that takes into account various factors influencing product valuations on Mercari, including brand, condition, and demand, to provide fair and competitive price suggestions.

The exploration of the dataset and its nuances is presented through an in-depth Exploratory Data Analysis (EDA), revealing crucial insights that guide subsequent model development. The process of data preprocessing is detailed to ensure that the dataset is cleaned and transformed for optimal model training.

The project proceeds with the introduction and evaluation of a benchmark solution, providing a baseline for performance comparison. Subsequent sections detail the iterative development of a first-cut solution and the exploration of deep learning-based techniques for addressing the price suggestion problem.

Approach 1- Machine Learning Models

Results

Results of ML Models

Approach 2- Deep Learning

Model Architecure

Kaggle Submission

Note: For detailed description check out the blog and Github Repository