Market Data Prediction Algorithm

I need an algorithm written in either C++ or Java that ingests live or historical market data and produces reliable forecasts of future price movements. This is a pure prediction task—no order execution logic or visual dashboards are required at this stage—so the focus is on building a robust core model that can process incoming numbers and output a clear price outlook. I will supply sample data feeds and a brief explanation of the columns once we start; your job is to design the prediction logic, implement it cleanly, and show me how to run it from the command line. Feel free to draw on statistical techniques, traditional machine-learning libraries, or your own handcrafted approach, as long as the solution is reproducible in C++ or Java and the code remains readable and well-commented. Deliverables • Fully-functional C++ or Java source code • Build/run instructions (e.g., Makefile, Maven, or Gradle) • A short README explaining the prediction methodology and any external libraries used • A sample run on my provided dataset demonstrating the output format Acceptance Criteria • Compiles without errors on a standard Linux or Windows setup • Consumes market data in CSV or JSON as provided and produces predicted prices for the next user-specified interval • Prediction accuracy metrics (MSE, MAE, or similar) printed after back-testing on the supplied sample set If you’ve built similar forecasting tools before, let me know; otherwise, show me a quick outline of your proposed approach when you bid.