Learning Outcomes:
On successful completion of this module, students should be able to:
1. Critically evaluate the principles, assumptions, strengths, and limitations of artificial intelligence, machine learning, deep learning, generative AI, and scientific large language models in scientific and engineering contexts;
2. Apply Python-based computational tools to acquire, clean, process, visualise, and analyse datasets relevant to agri-food, health, life science, and engineering applications;
3. Design and implement appropriate machine learning workflows, including dataset construction, feature engineering, model selection, training, validation, and performance evaluation;
4. Interpret and critically appraise model outputs, uncertainty, limitations, and potential sources of bias in relation to domain-specific scientific or engineering problems;
5. Apply and critically assess generative AI and large language model-based approaches for coding, data analysis, knowledge extraction, or scientific problem-solving, with consideration of reliability, reproducibility, and ethical use;
6. Design, execute, and communicate an AI-based project addressing a scientific or engineering problem, demonstrating appropriate methodological choices, integration of domain knowledge, and effective written and oral communication.
Indicative Module Content:
1. Introduction to AI in Science and Engineering
- Overview of artificial intelligence, machine learning, deep learning, generative AI, and scientific large language models.
- Applications of AI in agri-food, health, life science, biosystems, and engineering domains.
2. Python and Computational Environments
- Introduction to Python-based data analysis using tools such as Google Colab and VS Code.
- Use of key libraries for data handling and analysis, including pandas, NumPy, and data visualisation tools.
3. Data Preparation and Feature Engineering
- Dataset construction, data cleaning, missing value handling, and data splitting.
- Feature engineering, feature scaling, and preparation of scientific and engineering datasets for modelling.
4. Traditional Machine Learning Methods
- Introduction to supervised machine learning for regression and classification tasks.
- Model selection, training, validation, and interpretation of model outputs.
5. Model Evaluation and Optimisation
- Evaluation metrics for classification and regression, including accuracy, precision, recall, F1 score, RMSE, MAE, and R².
- Overfitting, underfitting, cross-validation, hyperparameter tuning, and reproducibility.
6. Introduction to Deep Learning
- Basic concepts of neural networks and deep learning.
- Applications to scientific and engineering data, including images, spectra, sensor data, biological sequences, and omics datasets.
7. Scientific Large Language Models
- Introduction to scientific and domain-specific large language models.
- Applications of LLM-based methods to molecular, biological, textual, and sequence-based data analysis.
8. Generative AI and AI-Assisted Coding
- Use of generative AI tools for coding support, debugging, data analysis, and rapid prototyping.
- Critical and responsible use of AI-assisted tools in scientific and engineering workflows.
9. Domain-Specific AI Case Studies
- Case studies in bioprocess optimisation, precision fermentation, bioactive compound discovery, food quality assessment, spectroscopy, sensor data analysis, formulation design, nutrition modelling, and biomedical data analysis.
- Critical discussion of how AI methods are selected, applied, evaluated, and interpreted in different application contexts.
10. AI Project Design and Communication
- Problem formulation, dataset selection, methodological justification, and workflow design for AI-based projects.
- Interpretation of results, discussion of limitations, presentation of findings, scientific writing, and acknowledgement of AI-assisted tools.