https://github.com/evilpegasus/real-estate-price-prediction
Predicting NYC real estate sale prices using neural networks (1st place Berkeley SAAS Kaggle Competition Fall 2020)
https://github.com/evilpegasus/real-estate-price-prediction
data ipynb kaggle nyc
Last synced: 8 months ago
JSON representation
Predicting NYC real estate sale prices using neural networks (1st place Berkeley SAAS Kaggle Competition Fall 2020)
- Host: GitHub
- URL: https://github.com/evilpegasus/real-estate-price-prediction
- Owner: evilpegasus
- Created: 2020-11-24T02:58:48.000Z (almost 5 years ago)
- Default Branch: main
- Last Pushed: 2020-12-18T02:40:28.000Z (almost 5 years ago)
- Last Synced: 2025-02-01T10:41:36.076Z (9 months ago)
- Topics: data, ipynb, kaggle, nyc
- Language: Jupyter Notebook
- Homepage: https://www.kaggle.com/c/saas-2020-fall-cx-kaggle-compeition/
- Size: 49.6 MB
- Stars: 8
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Real Estate Price Prediction
### Berkeley SAAS CX Fall 2020 Kaggle Competition
### Ming Fong and Yifan Zhang
**Winning solution**
Predicting real estate sale prices using property data.
The code used for our final submission can be found in [final_submission.ipynb](final_submission.ipynb).
## Data
Data can be downloaded from the [Kaggle competition data page](https://www.kaggle.com/c/saas-2020-fall-cx-kaggle-compeition/data).
In the repo, data is in the `/data` directory.
There are 3 data files:
* [data/test_features.csv](data/test_features.csv)
* [data/train_features.csv](data/train_features.csv)
* [data/train_targets.csv](data/train_features.csv)
[output/sample_submission.csv](output/sample_submission.csv) is an example of a file that is ready to submit to Kaggle. There are two columes: `id` and `SALE PRICE`.
## Kaggle Link
https://www.kaggle.com/c/saas-2020-fall-cx-kaggle-compeition
## Notes
Building codes: https://www1.nyc.gov/assets/finance/jump/hlpbldgcode.html
TODO
- remove outliers
- check negative price predicitons
-Check if building or tax class changes
- could mean redeveloped housing
- add column "classChanged" - 1 if yes, 0 if no
- Check if apartment number is present
- add column "hasApartmentNumber" - 1 or 0