This document discusses algorithmic trading using machine learning. It introduces algorithmic trading and its advantages over human trading. The objectives are to maximize profits with minimum capital and predict stock prices using machine learning. Python is used to develop the trading platform. The architecture includes exchanges, servers to receive and store market data, and applications to interface and manage orders. Strategies discussed include analyzing candlestick patterns and volume. A proposed algorithm uses machine learning on historical data to generate buy signals. Limitations include algorithms not being 100% accurate and not universal. Future enhancements may include greater use of artificial intelligence as technology advances.