Research map / Large language models / AI and society
Towards an AI Course Based on Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes and describes a proof-of-concept undergraduate AI course organized around neural networks, feature learning, and pattern recognition, with course materials and student feedback from a pilot.
Problem
AI curricula, following CS2023, give neural networks a back seat to symbol-based AI; the paper asks how to structure an introductory AI course around NN-based pattern recognition and feature learning.
Approach
Reframes an introductory AI course so that feature learning and pattern recognition are the foundation, with assignments implementing perceptrons, feed-forward nets, LeNet5-style CNNs and LLM fine-tuning, while still covering classical search and Deep Blue as contrast.
Key results
- Course is organized into learning outcomes, assessments (projects 50%, essays 10%, presentations 20%, quizzes 20%) and a sequence of NN-focused modules over a ten-week quarter.
- Students reported positive feedback on the refocused NN course, with the software assignments and readings cited as strengths.
- Following student feedback, the instructors revised FFnet and CNN assignments to be implemented in Java and added three quizzes.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| refocus the AI course on neural networks | proposed AI course · Spring term course evaluations | student-reported learning | Helped |
| include software development assignments | proposed AI course · Spring term course evaluations | student-reported learning | Helped |
| include readings | proposed AI course · Spring term course evaluations | student-reported learning | Helped |
Limitations
- Evaluation is subjective and based on anonymous student feedback from a single pilot course at one institution.
- Probability and LLM reasoning assignment were removed or made optional, and some assignments remain under development.
- The authors acknowledge the course is in flux and not perfect.
Experimental setup
Datasets: MNIST TinyStories Sudoku problems
Metrics: anonymous student course evaluations subjective instructor assessment
Builds on: CS2023 curriculum AI4K12 Five Big Ideas
Closest papers by meaning
- Computational Thinking with Computer Vision: Developing AI Competency in an Introductory Computer Science Course (AAAI 2025 · similarity 0.85)
- “AlphAI”: Teaching AI Algorithms to K12 by Training Learning Robots and Visualizing How It Works (AAAI 2025 · similarity 0.84)
- What Can Youth Learn About Artificial Intelligence and Machine Learning in One Hour? Examining How Hour of Code Activities Address the Five Big Ideas of AI (AAAI 2025 · similarity 0.83)
- Artificial Intelligence for Future Presidents: Teaching AI Literacy to Everyone (AAAI 2025 · similarity 0.83)
- Foundations of Autonomous Vehicles: A Curriculum Model for Developing Competencies in Artificial Intelligence and the Internet of Things for Grades 7–10 (AAAI 2024 · similarity 0.83)
- Comparing Artificial Intelligence Curricula in Canadian and US Universities (AAAI 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.