reSaleScout

Forecasting auction sale prices on different websites using random forest regression.

In this folder you will find notebooks, .py scripts, and supporting files used to run my application, reSaleScout, which uses historical sale data from two auction websites to predict where sellers will make a higher profit based on the type, brand, and condition of their women's clothing item.

Play around at: www.resale.fun (<- now defunct!)

Below, an overview of the contents of this repository:

eBay_scraper.ipynb, poshmark_scraper.ipynb

These notebooks detail my application of BeautifulSoup, pandas, NumPy, requests, etc. to scrape data for 10 of the most popular women's clothing brands listed on eBay and poshmark. The HTML and data storage differed between the two sites slightly, thus the two notebooks.

Data_exploration.ipynb

Cleaning and polishing some data.

Postgres_scrapers.ipynb

Setting up a PostgreSQL database and querying a bit.

Random forests.ipynb

Random forest instantiation- but that's not all! A hint of XGBoost as well.

/CSVs

Important .csv files I don't want to lose if my computer perishes.

.pkl, cat files

Pickled models, binary vectors, used in my application deployment scripts.

init.py, resale.py, run.py

The initialization, application, and executable files used to deploy my application on AWS.

landing.html

My template folder. There are files associated with this template that are replaceable- the guts of this HTML doc are the important bit.