COMPETITIONPredict The
Market
Put your skills to the test in this year's Kaggle competition! Level up and compete for prizes. Fit for all skill levels.
Competition starts in
14:17:25:31
DD · HH · MM · SS
COMPETITIONCOMPETITIONPut your skills to the test in this year's Kaggle competition! Level up and compete for prizes. Fit for all skill levels.
Competition starts in
14:17:25:31
DD · HH · MM · SS
Kaggle is the world's leading platform for data science and machine learning. It hosts real-world datasets, challenges and competitions where individuals and teams build solutions to complex problems. For our flagship competition, we source Kaggle datasets to give you practical, industry-relevant experience.
This year's challenge puts you in the driver's seat of the markets. Build models that forecast price movements, uncover signal in the noise, and compete on real historical data. Whether you lean on classic time-series methods, deep learning, or creative feature engineering — the goal is simple: predict the market better than anyone else in the room.
Sharpen your edge before the bell rings. Each session is recorded — catch up on the replay or grab the slides.
Dig into market data — clean it, visualise it, and surface the signals that actually move prices before you feed them into a model.
Recording drops after the session
Build, benchmark, and backtest candidate models. Learn how to evaluate predictions against real profit curves instead of just accuracy.
Recording drops after the session
The traders behind the desk — your education crew designs the curriculum, runs the workshops, and makes sure no one walks in cold.
Stanley heads up DSCubed Education — shaping the workshop syllabus, mentoring competitors, and making sure the Kaggle comp runs like a well-oiled order book.
Anyone — students, grads, and industry folk across all skill levels. You can compete solo or form a team of up to four. We have tracks and prizes designed to keep things fair whether it's your first Kaggle or your fiftieth.
Nope. Our workshops cover everything from exploratory data analysis to model building and evaluation. If you can write a bit of Python, you can ship a submission by the end of the event.
Submissions are ranked on a held-out test set using a combination of prediction score and simulated trading profit. The public leaderboard updates live; the final board locks when submissions close at 6pm on competition day.
A laptop, charger, and your brain. We supply datasets, starter notebooks, snacks, catered lunch, and enough coffee to keep you well past the closing bell.
Free for DSCubed members. Non-members can sign up at the door with a one-time membership fee that also gets you into every other event we run this year.
Hop into our Discord — there's a dedicated #kaggle-26 channel where organisers, mentors, and past competitors hang out and help out.