CSCI218 Foundations of Artificial Intelligence

Assignment

Due: Friday 04 Sept 2026, 11:30 PM Total marks: 15 marks

Overview

Automatic Pokémon image recognition has applications in augmented reality, gaming analytics, digital collectibles, and educational tools. In this assignment, you are required to implement Pokémon image classification using two types of classifiers: Multi-layer Perceptron (MLP) and Convolutional Neural Networks (CNN).

You are provided with a subset of a labelled Pokémon image dataset from Kaggle (https://www.kaggle.com/datasets/lantian773030/pokemonclassification), containing images categorized into 10 Pokémon species: Bulbasaur, Meowth, Mew, Pidgeot, Pikachu, Snorlax, Squirtle, Venusaur, Wartortle, and Zubat. Each category contains images of different sizes.

What you need to complete this assignment:

· The Pokémon image dataset, available on the Moodle site of this subject.

· Sample code provided on the Moodle site.

· A Python 3 programming environment with all required libraries and dependencies installed.

Note that:

· The sample code provides the essential functions for extracting features and loading the dataset. You must modify and complete the code to solve the tasks.

· You should maximally utilise the built-in functions in existing libraries, packages, and modules (e.g., scikit-learn) to complete this assignment.

· All tasks must be completed individually, without any collaboration with other students. If you have any questions or concerns, please consult your lecturer or tutor during lab sessions or consultation hours.

· The use of Generative AI tools (e.g., ChatGPT, Copilot, Gemini) is strictly prohibited for the assessment tasks. Any content identified as AI-generated will be treated as plagiarism and may result in a Fail grade for the subject.