Home
ES · Ver en español

Education project

Artificial Intelligence: From Rules to an Agent

Four ways for a program to decide, from an expert system to an agent

The course's four modules, with an expert system and a fuzzy one for candlestick charts, genetic algorithms, models for SmartPot and a reactive agent with its dashboard.

The agent's real Streamlit dashboard with lettuce selected, showing three anomalies detected and the actions it fires, such as turning on the fans and the water cooler and diluting the solution. (open the full-size image)
The lower part of the real dashboard, with the crop's status, the agent's actions and temperature and humidity against their ideal range. (open the full-size image)

What it is about

The Artificial Intelligence repository holds the course's four quizzes, each with its own paradigm, which are symbolic AI, evolutionary computing, supervised learning and agents. I did the first two in a pair.

The last two use SmartPot as the case, with models that decide when to ventilate or correct pH and a reactive agent that looks at the current sensor reading and answers with actions, without memory of earlier readings.

What it does

  • An expert system in CLIPS, called from Python with clipspy and with a Tkinter interface, that recognizes bullish and bearish engulfing patterns in two candlesticks.
  • A fuzzy logic system with scikit-fuzzy that uses the candle sizes and the engulfing difference to recommend investing a little, a normal amount or a lot.
  • Genetic algorithms that rebuild an image's color from its grayscale version, and that optimize a skydiver's jump to land on a given point.
  • For SmartPot, a logistic regression that controls the fans and a neural network that doses pH, with desktop and Streamlit versions.
  • A reactive agent with a Streamlit dashboard that compares each reading with the crop's ranges, tomato or lettuce, and decides which actuators to turn on.

Decisions and their costs

  1. Rules in a separate knowledge base

    If each crop's ranges lived in the agent's logic, adding a crop would mean changing it.

  2. A dashboard to test the agent

    Testing the agent required the sensors to be connected.

Repositories

Every public repository of the project, with what it holds.

  • 2025

    IAEngineeringSebastian-s

    Artificial intelligence in four modules, with a CLIPS expert system, fuzzy logic, genetic algorithms that colorize images and classification models for SmartPot.

    Python, CLIPS, scikit-fuzzy, scikit-learn
GitHub organization: EngineeringSebastian-s

Stack

Pick a technology to see the projects that use it.

Related