A colorful browser-based Rock Paper Scissors game powered by AI and machine learning.

The game uses Phaser for the interactive game interface and
TensorFlow.js to run machine learning directly in the browser.
The AI observes the player’s previous moves and learns their playing patterns in real time.

After enough rounds, the machine-learning model becomes active and starts predicting
the player’s next move. The computer then chooses a counter-move based on the prediction.

Key Features

• 🎮 Browser-based game — runs directly in a modern web browser
• 🤖 AI-powered opponent — the computer learns from the player’s behavior
• 🧠 Machine learning — uses a small neural network running with TensorFlow.js
• ⚡ Client-side AI — all machine-learning calculations run locally in the browser
• 🔒 Privacy-friendly — no player data needs to be sent to a server
• 🎨 Colorful visual interface — illustrated Rock, Paper, and Scissors icons
• 📊 Game statistics — displays player wins, computer wins, draws, and win rates
• 🔮 Prediction display — shows what move the AI currently predicts
• 📈 Learning status — clearly indicates when the machine-learning system becomes active
• 🕹️ Phaser framework — handles the interactive game experience

Technology

JavaScript · Phaser · TensorFlow.js · HTML5 · CSS

The project demonstrates how AI and machine learning can be integrated
into a lightweight browser game without requiring a backend server.