/sneaker-forecast-app

A price forecaster for the sneaker resale market

Primary LanguageJupyter Notebook

sneaker-forecast-app

An app that predicts sneaker resale prices in the near-future.

screenshot

A broad overview of this project:

The objectives are:

  1. Predict prices on the resale market for the next 7 days
  2. Collect important information about sneakers that can help inform buying or selling decisions

General tools and method can be broken into 4 parts:

1. Data collection

  • Combination of scrapy and selenium to collect information for this project
  • Pandas to clean and transform data for modelling

2. Linear regression

  • scikit-learn for price prediction

3. NLP

  • Word2Vec for processing social media specific words
  • Gensim for sentiment analysis

4. Web app

  • Flask/javascript used to communicate results in the web app
  • HTML/CSS/Bootstrap for styling