Rocket Capital Investment (RCI), in association with MachineHack, is set to launch the longest blockchain tournament from June 13 to September 05 — a twelve-week hackathon to source and incentivise the best in machine learning applications for finance. The tournament is made up of a series of challenges, where the participants stand a chance to win over INR 4,00,000 cash & 200,000+ MUSA coins.
| Note: The MUSA coin is the first token fully dedicated to the data scientist community. Read more here. |
Being a Licensed Financial Institution headquartered in Singapore, RCI combines their financial expertise with external machine learning forecasts through a blockchain tournament on financial markets.
The company believes that a key need for RCI is to feed its engine with thousands of machine learning predictions from data scientists to improve the accuracy of the Meta-Model Rocket used to make portfolio asset allocation.
Through this competition, RCI aims to use a decentralised platform to source and incentivise the best in machine-learning applications for the finance industry. Alongside, data scientists and ML practitioners can now be rewarded weekly with MUSA coins for their data science skills. This can also help them create a decentralised track record of their performance.
Start Date: June 13, 2022
End Date: September 05, 2022
Register Now
All about the #CryptoPrediction challenge
In this challenge, RCI asks the data scientists & ML Practitioners to submit financial predictions using the weekly dataset. RCI evaluates participants’ performance and pays them rewards at the end of the week.
Let’s not forget that the MUSA token drives the challenge, where participants stake their MUSA coins to submit their predictions and earn rewards based on their stake and the performance of their submissions. The community competition gets up to 200,000 tokens and a 500 USDT prize (INR 40,000) at the end of every week, for the next 12 weeks.
How does it work?
The #CryptoPrediction tournament is made up of weekly challenges, where each week:
- RCI distributes a dataset (every Monday at 11:00 AM, till September 05) and opens the submission window.
- Participants can start submitting their crypto predictions of the financial markets.
- At the end of the week, RCI closes the submission window (every Tuesday at 5:30 AM, till September 05).
- Post that, the predictions are evaluated against the real-world data, and the rewards are paid out in MUSA tokens & cash amount.
Dataset
The dataset consists of
- #first column (symbol) is the ticker
- # ‘target_DC1’ is the log of return for the delta between the current close to the previous close
- # ‘target’ is the y which is ranked by Era using ‘target_DC1.’
- # the rest of the columns are features to be used for ML
Start Date: 13th June 2022
End Date: 05th September 2022
Register Now
Prerequisites for the #CryptoPrediction challenge
- Follow https://twitter.com/rocketcapinv on Twitter
- Join the RocketCapital Discord channel at https://discord.com/invite/r6NbnCSrzn
- Open an ETH wallet and configure tournament dApp ( https://www.youtube.com/watch?v=Lb_A8J2wJMw&t=1s)
- Request MUSA airdrop on this site ( https://rocketcapital.ai/airdrop/)
- Check for airdropped MUSA tokens on your wallet or competition account
- Use MUSA tokens for staking ( https://www.youtube.com/watch?v=5CLKh776mtI&t=50s)
- Each week download the dataset and send the predictions ( https://www.youtube.com/watch?v=A9iUuV1RVdo )
- The challenge will only allow for one submission each week per wallet address (One Account Per Wallet Address), and submissions from multiple accounts will lead to disqualification. All registered users are eligible to participate in the hackathon.
- Privately sharing code or data outside of teams is not permitted.
- Rewards will be paid to the wallet address used to submit, so please coordinate with your team regarding the distribution of rewards.
- Each team can have a maximum of 5 members aged above 18. The one who shared the request (once accepted by another participant) will become a team leader.
- Only a team leader can create a team (including sharing requests & accepting members). A team or an individual participant will be treated as one entity. A single participant adding a submission will be discarded once they join a team.
- Changes to team members are not permitted. Once the team is formed, it is not possible to delete the team.
How will you be evaluated?
The submission will be evaluated using the Spearman Correlation metric. One can use ‘scipy.stats.spearmanr(prediction, target)’ to calculate the same. [mean_squared_error(y_true, y_pred, squared=False)]
The Final Score represents the score achieved based on the Best Score on the public leaderboard.
Rewards will be given to participants according to the above scores, considering the correlation among predictions (refer to RCI Competition for technical details).
Sklearn models support the predict() method to generate the predicted values. The participant should submit a .csv file with the same number of rows of the test dataset. The submission will return an Invalid Score if you have extra rows or columns. The file should have exactly two columns (ticker and predicted value).
PLEASE NOTE:
- Do not shuffle the sequence of the test series.
- If you are using pandas, use this submission code-submission_df.to_csv(‘my_submission_file.csv’, index=False)
To stake MUSA coins and make submissions, participants will need a Metamask Wallet. This will help you manage your tokens and take action on the blockchain.
