https://github.com/maximilien/auction-ai
An ML / AI project for predicting auction prices of goods
https://github.com/maximilien/auction-ai
Last synced: 3 months ago
JSON representation
An ML / AI project for predicting auction prices of goods
- Host: GitHub
- URL: https://github.com/maximilien/auction-ai
- Owner: maximilien
- License: apache-2.0
- Created: 2024-09-16T21:49:02.000Z (8 months ago)
- Default Branch: main
- Last Pushed: 2025-01-13T23:49:05.000Z (4 months ago)
- Last Synced: 2025-01-14T00:27:26.159Z (4 months ago)
- Language: Jupyter Notebook
- Size: 4.03 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
### Exploring Luxury Auction
**E. M. Maximilien**
#### Executive summary
In this project we are exploring luxury auction data and attempt to answer the question:
`What are the key factors that help increase selling prices of luxury auction items?`
#### Rationale
During the pandemic, the luxury industry saw a radid increase in luxury purchases and collecting, e.g., wristwatches. While this lasted a few years, the rapid decline in interst and increase in prices (inflation) caused a reverse. These twin facts imply that there will be an increase of luxury items on the market and at auctions houses.
#### Research Question
What are the key factors that help increase selling prices of luxury auction items?
#### Data Sources
[CSV](luxury_auction.csv) file from [Kaggle](https://www.kaggle.com/datasets/onlineauctions/online-auctions-dataset)
#### Methodology
* Use basic explorary data analysis
* Build regression and classifier models
* Hyperparameter tuning of models
#### Results
* `SVC` had the best accuracy to predict which item would sell and price but took long time to improve
* `LinearRegression` was good enough with accuracy score close to the best models
* Most important features were `bidder_rate` and rhe `price`
* The `auction_type` also helped determine the `price` and likely the `bidder_rate`
* The `bidder_rate` increase correlated with `price` increase
* The `bidder` had little impact on resulting `price`
#### Next steps
* Explore what features we could discard
* Explore the most important features in the best models
* Explore whether bid time impact sell price
* Explore if bidder had impact on sell price
#### Outline of project
- [Link to notebook](auction-ai.ipynb)
##### Contact and Further Information
[mmaximilien at Gmail](mailto:[email protected])