An applied Swiss-inspired curriculum combining algorithmic rigour, systems architecture, digital trust and practical engineering capability for modern technology leaders.
The BSc in Computer Science with Artificial Intelligence at Brookfield School of Computing is engineered to provide deep competency across core computer science, systems design, and targeted domain applications. Students develop fluency in software craftsmanship, distributed architectures, algorithm optimization, and data governance.
Instruction balances rigorous theoretical foundations with practical code repositories, laboratory simulations, and ethical technology review aligned with modern international computing standards.
| Code | Module Title | Type | Discipline & Learning Outcome |
|---|---|---|---|
| AIM101 | Foundations of Artificial Intelligence & Rational Agent Architectures | Theory | Formulate rational agent environments, PEAS architectures and state-space representations for intelligent Artificial Intelligence systems. |
| AIM102 | Scientific Python & Tensor Computing Lab (NumPy, SciPy) | Practical | Vectorize multidimensional tensor computations and matrix operations using NumPy, SciPy and scientific libraries. |
| AIM103 | Linear Algebra, Vector Spaces & Matrix Decomposition for AI | Theory | Compute eigenvalues, eigenvectors, SVD and orthogonal projections for multidimensional feature representations. |
| AIM104 | Heuristic Search, A* & State-Space Exploration Lab | Practical | Implement A* search, minimax with alpha-beta pruning, constraint satisfaction and genetic algorithms in Python. |
| AIM105 | Probability Theory, Random Variables & Bayesian Inference | Theory | Apply Bayes' theorem, probability density functions, maximum likelihood estimation and probabilistic reasoning. |
| AIM106 | Relational & Vector Databases for Embedding Retrieval Lab | Practical | Deploy PostgreSQL with pgvector to build similarity search indices and nearest-neighbor vector queries. |
| AIM107 | Knowledge Representation, Ontologies & Description Logics | Theory | Model domain knowledge graphs, semantic ontologies and first-order predicate logic inference engines. |
| AIM108 | Data Acquisition, Preprocessing & Feature Engineering Lab | Practical | Automate data extraction, missing value imputation, categorical encoding and feature scaling pipelines. |
| AIM109 | Ethics of Autonomous Systems, Algorithmic Bias & AI Safety | Theory | Evaluate algorithmic fairness metrics, societal impacts of AI decisions and ethical safety guardrails in Artificial Intelligence. |
| Code | Module Title | Type | Discipline & Learning Outcome |
|---|---|---|---|
| AIM201 | Statistical Machine Learning & Supervised Classification | Theory | Derive and evaluate decision trees, support vector machines, logistic regression and naive Bayes classifiers. |
| AIM202 | Supervised & Unsupervised Learning Implementation Lab | Practical | Train, tune and validate cross-validated predictive models using Scikit-Learn and ensemble techniques. |
| AIM203 | Multivariable Calculus & Gradient Optimization for AI | Theory | Compute gradient descent trajectories, Jacobian matrices, Hessians and Lagrange multipliers for loss surfaces. |
| AIM204 | Neural Networks Fundamentals & Deep Learning with PyTorch Lab | Practical | Construct feedforward neural networks, backpropagation algorithms and custom PyTorch loss modules. |
| AIM205 | Computer Vision: Feature Extraction & Convolutional Networks | Theory | Design 2D convolutional layers, pooling mechanisms, stride filters and spatial hierarchies for image understanding. |
| AIM206 | Natural Language Processing: Text Embeddings & Tokenization Lab | Practical | Build n-gram models, Word2Vec embeddings, tokenizers and sentiment classification pipelines in Python. |
| AIM207 | Reinforcement Learning Fundamentals & Markov Decision Processes | Theory | Formulate Bellman optimality equations, value iteration and policy iteration in dynamic MDP environments. |
| AIM208 | High-Performance GPU Acceleration & CUDA Computing Workshop | Practical | Optimize batch tensor parallel processing on NVIDIA CUDA GPU clusters and cloud acceleration platforms. |
| AIM209 | Applied Artificial Intelligence First-Year Project: Predictive Model Build | Practical | Deliver an operational end-to-end predictive machine learning model integrated with REST API endpoints. |
| Code | Module Title | Type | Discipline & Learning Outcome |
|---|---|---|---|
| AIM301 | Deep Neural Network Architectures & Optimization Dynamics | Theory | Analyze vanishing gradients, batch normalization, Adam/RMSprop optimizers and residual skip connections. |
| AIM302 | Transformer Architectures & Self-Attention Implementation Lab | Practical | Implement multi-head self-attention mechanisms and transformer encoder blocks from scratch in PyTorch. |
| AIM303 | Distributed Deep Learning & Multi-GPU Training Systems | Theory | Formulate data-parallel (DDP) and model-parallel distributed training across cluster computing nodes. |
| AIM304 | Containerized AI Environments with Docker & GPU Drivers Lab | Practical | Package reproducible machine learning environments utilizing NVIDIA-Docker containers and CUDA runtimes. |
| AIM305 | Graph Neural Networks & Relational Knowledge Learning | Theory | Model complex relational networks using message passing neural networks (MPNN) and graph convolutional layers. |
| AIM306 | Automated Model Validation, Hyperparameter Tuning & Optuna Lab | Practical | Automate Bayesian hyperparameter searches, pruning algorithms and cross-validation pipelines with Optuna. |
| AIM307 | Explainable AI (XAI), Feature Attribution & SHAP/LIME | Theory | Evaluate black-box neural predictions using SHAP values, integrated gradients and LIME local surrogates. |
| AIM308 | Adversarial Machine Learning & Robustness Testing Lab | Practical | Generate FGSM adversarial perturbations, defend against poisoning attacks and measure model robustness bounds. |
| AIM309 | Empirical Research Methods in Machine Intelligence & Artificial Intelligence | Theory | Design rigorous statistical hypothesis tests, benchmark datasets and ablation studies for AI research. |
| Code | Module Title | Type | Discipline & Learning Outcome |
|---|---|---|---|
| AIM401 | Generative AI, Latent Diffusion & Variational Autoencoders | Theory | Formulate variational autoencoder (VAE) evidence lower bounds and denoising diffusion probabilistic models (DDPM). |
| AIM402 | Large Language Model Fine-Tuning & LoRA / QLoRA Lab | Practical | Fine-tune open-source LLMs using parameter-efficient fine-tuning (PEFT), LoRA and quantization techniques. |
| AIM403 | MLOps Principles, Feature Stores & Continuous Training | Theory | Architect enterprise MLOps lifecycles connecting feature stores, model registries and continuous retraining. |
| AIM404 | MLflow, DVC & Kubeflow Pipeline Engineering Lab | Practical | Deploy automated reproducible training pipelines using MLflow tracking, DVC data versioning and Kubeflow. |
| AIM405 | AI Safety, Value Alignment & Hallucination Mitigation | Theory | Evaluate reinforcement learning from human feedback (RLHF), constitutional AI and output guardrail systems. |
| AIM406 | CUDA C++ Kernel Programming & Hardware Tensor Acceleration Lab | Practical | Write custom GPU CUDA C++ kernels, shared memory optimizations and warp-level reduction primitives. |
| AIM407 | Regulatory AI Compliance, EU AI Act & Swiss Data Governance | Theory | Audit high-risk AI models against the EU AI Act conformity requirements, Swiss FADP and transparency mandates. |
| AIM408 | Generative Image & Multimodal Diffusion Implementation Lab | Practical | Build custom Stable Diffusion conditioning pipelines, ControlNet adapters and CLIP image-text embeddings. |
| AIM409 | Industrial Artificial Intelligence Team Project: System Engineering Studio | Practical | Engineer and deploy an enterprise-grade AI system with automated telemetry, CI/CD and production endpoints. |
| Code | Module Title | Type | Discipline & Learning Outcome |
|---|---|---|---|
| AIM501 | Multimodal Foundation Models & Cross-Attention Systems | Theory | Deconstruct vision-language models (VLM), cross-attention fusion and embodied multimodal perception. |
| AIM502 | Triton Inference Server Deployment & High-Throughput Serving Lab | Practical | Deploy optimized ONNX and TensorRT model pipelines onto NVIDIA Triton Inference Server with dynamic batching. |
| AIM503 | Autonomous Multi-Agent Swarms & Hierarchical Planning | Theory | Design decentralized multi-agent coordination, auction protocols and hierarchical reinforcement planning. |
| AIM504 | Advanced Deep Specialization Studio in Artificial Intelligence | Practical | Implement cutting-edge domain neural architectures from recent peer-reviewed scientific literature in Artificial Intelligence. |
| AIM505 | Enterprise AI Strategy, ROI & Technological Innovation | Theory | Formulate corporate AI adoption roadmaps, evaluate compute cost ROI and architect enterprise data assets. |
| AIM506 | Model Inversion & Membership Inference Defense Lab | Practical | Simulate model extraction attacks and deploy differential privacy noise mechanisms to safeguard training data. |
| AIM507 | Swiss & Global Standards for Trustworthy & Auditable AI | Theory | Apply ISO/IEC 42001 AI management standards, algorithmic audit frameworks and technical documentation norms. |
| AIM508 | Model Drift, Data Shift & Evidentiary Telemetry Lab | Practical | Configure Evidently AI and Prometheus monitors to detect covariate shift, concept drift and performance decay. |
| AIM509 | Senior Artificial Intelligence Capstone Project: Architecture Specification | Practical | Author comprehensive technical specifications, model card blueprints and empirical validation protocols. |
| Code | Module Title | Type | Discipline & Learning Outcome |
|---|---|---|---|
| AIM601 | Artificial General Intelligence Horizons & Neuro-Symbolic AI | Theory | Evaluate neuro-symbolic reasoning, common sense knowledge graphs and theoretical bounds of artificial general intelligence. |
| AIM602 | Senior Artificial Intelligence Capstone Project: Production Pipeline Build | Practical | Train, optimize and package the complete capstone AI model pipeline into production-grade infrastructure. |
| AIM603 | Senior Artificial Intelligence Capstone Project: Empirical Benchmarking & Ablation | Practical | Execute rigorous ablation studies, benchmark against baseline state-of-the-art models and validate error bounds. |
| AIM604 | Strategic AI Leadership & Boardroom Communication | Theory | Translate complex algorithmic tradeoffs, accuracy-latency budgets and ethical considerations to executive leadership. |
| AIM605 | Model Quantization, Pruning & Edge Deployment Lab (TensorRT) | Practical | Quantize neural weights to INT8/FP4, apply structured pruning and benchmark inference on edge devices. |
| AIM606 | Societal Impact, Labor Economics & Philosophy of Mind in AI | Theory | Analyze labor market transformation, consciousness theories, intellectual property of generative outputs and ethics. |
| AIM607 | Industry Technical Portfolio & Model Card Defense | Practical | Defend model architectures, hyperparameter choices, fairness metrics and carbon footprint before an expert panel. |
| AIM608 | Peer Review, Scientific Dissemination & Technology Transfer | Theory | Author peer-reviewable research papers, reproducibility checklists and open-source benchmark repositories. |
| AIM609 | Final Artificial Intelligence Capstone Project Demonstration & Oral Defense | Practical | Demonstrate the live production AI system and successfully pass an oral academic defense examination in Artificial Intelligence. |
Secondary School Diploma / Swiss Matura / International Baccalaureate or recognized foreign upper-secondary certificate. English proficiency IELTS 6.0+ or equivalent.
Brookfield School of Computing is an independent private education institution in formation. Academic programmes, entry requirements, fees and awards are subject to institutional review and applicable Swiss legal regulations.