Zindi International Competition to Detect Electricity and Gas Consumption Fraud
Applicant criteria
- NationalityNo specific nationality required
- AgeNo specific age required
- GenderBoth
Opportunity criteria
- Job locationTunisia
- Deadlineavailable
Opportunity description
Zindi hosts the Fraud Detection in Electricity and Gas Consumption Challenge, inviting participants worldwide to build models that help Tunisia’s STEG identify customers involved in meter fraud. This ongoing learning competition uses customer billing history and does not have a closing date.
About the Fraud Detection Competition
The challenge focuses on detecting and recognizing customers involved in fraudulent activity by analyzing their billing history. Its host, the Tunisian Company of Electricity and Gas (STEG), supplies electricity and gas across Tunisia. The competition page says meter manipulation has caused the company losses of around 200 million Tunisian dinars, and that solutions could help increase revenue and reduce losses.
Participants work on a classification problem. The evaluation metric is Area Under the Curve (AUC). Submission files should contain a client_id column and a target column, with the values formatted as shown in the competition’s sample. The page includes a tutorial for the challenge and links to general Zindi getting-started resources, including guidance on running notebooks in Colab, creating a team and updating a profile.
This is a learning competition, not a cash-prize contest. Participants receive 25 Zindi points for their first submission and 50 points for their first non-sample submission. The competition does not close, although Zindi reserves the right to update its timeline. Participants can submit up to 10 times per day. A submission that fails during processing is not counted toward that daily limit, and there is no public/private leaderboard split.
Who Should Apply
The challenge is open to participants in any country who want to practice classification and use data analysis to identify possible fraud patterns in billing records. You may participate individually or form a team of up to four people. Anyone taking part must follow the competition’s rules on data use, submissions and collaboration.
Key Highlights
- Challenge: detect fraudulent activity using customer billing history.
- Host: Zindi; challenge host: STEG in Tunisia.
- Evaluation metric: Area Under the Curve (AUC).
- Open to participants worldwide; individual and team participation are allowed.
- Ongoing competition with no closing date and no prizes beyond Zindi points.
Explore other competitions and awards or browse relevant AI opportunities. You can also find more opportunities for youth across different fields.
Benefits
- Receive 25 Zindi points for the first submission.
- Receive 50 Zindi points for the first non-sample submission.
- Gain knowledge through a learning challenge focused on classification and fraud detection.
Eligibility criteria
- The challenge is open to participants from all countries.
- Participation may be individual or in a team of up to four people.
- A team’s total submissions must be no greater than the maximum allowed at the time it is formed. The team receives the competition’s maximum submission allowance minus the highest number of submissions held by any team member when the team is formed.
- Multiple accounts per user are not permitted. Participants may not collaborate or be members across multiple teams.
- Code must not be shared privately outside a team. Any shared code must be made available to all competition participants through the platform, such as on discussion boards.
- Use publicly available open-source packages and only the datasets provided for the competition. The provided metadata and automated machine learning tools such as AutoML may not be used.
- Pretrained models may be used if they are openly available to everyone.
- Competition data must remain private and must not be transmitted, published or redistributed to non-participants. Cloud platforms such as Google Colab or AWS may be used if the data remains private and Zindi’s rules are followed.
Ready to apply?
Zindi International Competition to Detect Electricity and Gas Consumption Fraud