RL Dojo

Strategy CompareBlackjack DemoLogin

Reinforcement Learning Practice Gym

Reinforcement Learning Dojo

RL Dojo is a hands-on practice gym for sequential decision making, where each round helps you improve policy-level thinking.

What Is RL Dojo?

RL Dojo is a practical training space for reinforcement learning habits: observe state, choose action, read reward, and iterate. Start with guided blackjack scenarios and grow toward broader decision experiments.

Watch AI Demo

Live Decision Engine

Probability updates after every dealt card.

Explainable Moves

Every suggestion includes short strategy logic.

Risk Modes

Conservative, balanced, and aggressive play styles.

01

Scan Table

Track player cards, dealer card, and remaining deck signals.

02

Score Outcomes

AI estimates expected value for every legal decision.

03

Guide Action

Receive one best move with confidence and short rationale.

Latest Blog Posts

See all

JayJayOct 6, 2026

Mathematically explaining SVD.

SVD is a method for decomposing the data matrix X into U, Sigma, and V. V is created from the eigenvectors of X transpose X, and Sigma is formed from the square roots of the eigenvalues. Then, U is obtained using the relationship X v = sigma u. Ultimately, X = U Sigma V transpose.

JayJayOct 3, 2026

Why is PCA necessary when there is a lot of data? — From Covariance to Eigenvector, Eigenvalue, and SVD.

Summary: As the number of features increases, calculations become more complex, and similar information may be duplicated. PCA reduces dimensions by finding the important directions of the data while preserving as much information as possible, rather than simply deleting features. To achieve this, it identifies relationships between features using covariance and finds important directions through eigenvectors and eigenvalues. Finally, we will examine a practical calculation example and its relationship with SVD.

This is the latest post page integrated into RLDojo.

See all