GENRunner

An application for managing experiments with multi-objective evolutionary algorithms (MOEA) based on the MOEA Framework. The system operates in a client-server architecture, where:

• Server – asynchronously runs experiments, collects results, and manages statistics.
• Client – a console application that allows launching experiments, viewing statuses and results, and managing experiments.

Features

• Running experiments:

  • Utilizes the MOEA Framework.
  • Experiment parameters: algorithm names, test problems, metrics, and budget (number of objective function calls or iterations).
  • Configuring multiple algorithms and problems within a single experiment.
  • Experiment repetitions (running an experiment in multiple instances).

• Asynchronous experiment execution:

  • Running multiple experiments simultaneously, considering machine constraints.

• Experiment management:

  • Listing ongoing and completed experiments.
  • Checking experiment status (running, completed, error).
  • Viewing results – tabular presentation (iterations as rows, metrics as columns).
  • Aggregating statistics (mean, median, standard deviation) with date range filtering.
  • Filtering experiment instances by algorithm, problem, or metrics.

• Results presentation:

  • Exporting results to a CSV file.
  • Generating graphs (PNG) – metric value charts over iterations.
  • New presentation formats compatible with existing mechanisms (e.g., time-based filtering).

• Experiment group management:

  • Creating experiment groups by tagging (by ID or filtered results) and adding to existing groups.
  • Displaying results for experiments within a specific group.

• API handling:

  • The CLI client communicates with the server, handles errors, and transmits experiment parameters.

• Experiment deletion:

  • Deleting individual experiments or entire groups (including results and metadata).

Team:

• Jakub Worek
• Jakub Białecki
• Bartłomiej Treśka

How to Run:

• Start MongoDB in Docker (instructions in the database folder).
• Run backend/src/main/java/pl/edu/agh/GenRunnerApplication.java.
• Run cli/src/main/java/cli/Client.java.