Task-Based Learning Opportunities in AI

In these activities, you’ll get hands-on experience by using pre-built AI models and toolkits to create simple AI applications. You’ll work on small projects that match your classroom setting, learning by doing!

• Learning Objectives:

  • Understand basic AI concepts by using existing models.
  • Follow a step-by-step process: load a dataset, train a model, and interpret the results.
  • Gain practical problem-solving skills through real-world projects.

• Example Activity: Mini AI Tasks

  • What You’ll Do:
    • Use a pre-trained model (like a facial recognition or text classification model) from libraries such as scikit-learn, TensorFlow, or PyTorch.
    • Test the model with your own data to see how it works.
  • Step-by-Step Guide:
    • Dataset Load: Follow a checklist to load your dataset properly.
    • Model Training: Run the model using the provided tools.
    • Result Interpretation: Analyze the output using simple metrics like accuracy or recall.
  • Sharing and Discussion:
    • Present your results to the class.
    • Discuss how well your model worked and brainstorm ideas to improve it.

Mastery of Advanced Skills in AI

Once you’re comfortable using simple models, you’ll move on to more advanced skills. This means learning how to decide when AI is the right tool for a job, reviewing data requirements, and even tweaking model settings.

• Learning Objectives:

  • Evaluate whether AI is appropriate for a given task.
  • Understand data collection and processing needs on your own.
  • Learn to choose between low-code tools and full-code programming based on the situation.
  • Develop skills to fine-tune models by adjusting parameters like hyperparameters and network architecture.

• Example Activity: Technology Suitability Debate

  • What You’ll Do:
    • Examine various AI use cases, such as medical diagnosis, image filtering, or chatbot development.
    • Discuss whether AI is truly necessary for these tasks and if the data available is sufficient and high quality.
  • Debate Points:
    • Is AI the best tool for this problem?
    • What might be the challenges with the current data?
    • What improvements could be made to the model or process?

• Example Activity: Hands-on Coding Level Selection

  • What You’ll Do:
    • Complete the same project twice: once using a low-code tool (like App Inventor or Power Apps) and once using full-code solutions (like Python libraries).
    • Compare the advantages and disadvantages of each approach.
  • Outcome:
    • Understand when a low-code approach might be faster and easier, and when full-code gives you more control over your project.

• Example Activity: Hyperparameter Tuning Practice

  • What You’ll Do:
    • Train an image classification model and adjust settings like epochs, learning rate, and batch size.
    • Record how these changes affect the model’s accuracy.
  • Outcome:
    • Learn how fine-tuning can improve your model and find the best settings for optimal performance.

Key Takeaways

• Practical, Hands-On Learning:

  • By using pre-trained models and toolkits, you’ll learn AI concepts by doing.
  • Follow a clear process (data load, model training, result interpretation) to build your skills.

• Advanced Skills for Real-World Problems:

  • Decide when to use AI and learn how to collect and process the necessary data.
  • Practice both low-code and full-code approaches, and learn to adjust model parameters to optimize performance.

• Active Discussion and Collaboration:

  • Engage in debates and group projects to deepen your understanding.
  • Share your findings and work together to improve your AI projects.

These activities will help you build a strong foundation in AI while also preparing you for future challenges in technology. Whether you’re creating a simple project or fine-tuning an advanced model, you’ll gain valuable skills that will be useful in school and beyond!

0