# **Competition Rules**

To ensure a fair and engaging competition, please adhere to the following rules. These guidelines are set to create a level playing field for all participants while encouraging creativity and transparency.

## **General Rules**
1. **Data Usage:**
   - Participants are encouraged to use external data sources to enrich their models. Any additional data used must be publicly available or shared openly with a source link.
   - Do not attempt to re-identify anonymized data or use private datasets that cannot be shared with the broader community.

2. **Submissions:**
   - Each team is allowed a maximum of 20 submissions per day.
   - Your final submission should be a CSV file following the format outlined in the competition details. The file should contain the columns `hex_id` and `cost_of_living`.
   
3. **Code and Model Requirements:**
   - Participants will be asked to show the GitHub repository with all the code used for the competition when called to interview. The code should be in Python, well-organized.
   - The repository should be public or accessible to the competition organizers. It must include a `README.md` file that details your methodology, data sources, and any feature engineering steps.
   - Ensure that your code does not include any malicious elements.

## **Evaluation Criteria**
1. **Primary Metric:** The primary evaluation metric is **Mean Absolute Error (MAE)**, which will measure the accuracy of your predictions for `cost_of_living`.
2. **Secondary Metrics:**
   - **Feature Engineering**: The creativity and effectiveness of feature engineering will be closely examined. This includes the use of external variables, meaningful transformations, and feature selection.
   - **Predictive Power**: We will assess the strength of the features you create in predicting the target variable.
   - **Code Quality**: Your code will be reviewed for clarity, organization, efficiency, and reproducibility.
   

## **Leaderboard Rules**
1. The public leaderboard will be based on 50% of the test data, with the remaining 50% reserved for the final evaluation on the private leaderboard.
2. The leaderboard rankings during the competition are provisional and may change based on the final evaluation using the private test set.

## **Intellectual Property**
1. By participating in the competition, you agree that the code, models, and data processing techniques you submit may be reviewed and potentially shared as learning material or examples.

## **Prohibited Actions**
1. Do not engage in any form of **collusion**, including sharing code or predictions between teams.
2. **Cheating** or using unfair methods to manipulate the leaderboard, such as exploiting bugs, is strictly prohibited.
3. Avoid any form of **reverse engineering** of the dataset to access private or sensitive information.
4. Any attempt to tamper with the competition platform, submission system, or evaluation process will result in disqualification.


## **Support and Questions**
- A discussion forum will be available for participants to ask questions, share insights, and discuss the competition.
- The competition organizers will monitor the forum and provide clarifications as needed. However, organizers will not provide direct assistance with code or models.
- If you have any concerns or suspect rule violations, please report them directly to the organizers through the provided contact channels.

## **Final Decision**
The competition organizers' decisions are final and binding in all matters related to the competition, including rule clarifications and dispute resolutions.

## **Good Luck!**
We wish all participants the best of luck!