TUTOR

M05 · Level I — Computational Intelligence Foundations

Scientific Programming & Software Engineering for Intelligent Systems

Understand how computational ideas become reliable, efficient and reproducible software systems.

Topics

  • First-Principles Computing
  • Memory
  • Numerical Software
  • Testing
  • Profiling
  • Reproducibility
  • Systems Engineering

Technical concepts

  • 01First Principle: What Is a Program?
  • 02Binary, Bytes & Machine Representation
  • 03Types, Semantics & Contracts
  • 04Functions, State & Side Effects
  • 05Memory: Stack, Heap, Lifetime & Ownership
  • 06Arrays, Strides & Contiguous Data
  • 07Vectorisation: From Loops to Array Operations
  • 08Numerical Reproducibility in Code
  • 09Testing as an Executable Specification
  • 10Assertions, Invariants & Defensive Programming
  • 11Debugging: Hypothesis Before Patch
  • 12Git as a Scientific Instrument
  • 13Environments, Dependencies & Containers
  • 14Profiling: Measure Before Optimising
  • 15Big-O vs Real Performance
  • 16Benchmark Design
  • 17Concurrency, Parallelism & Determinism
  • 18APIs, Interfaces & Data Contracts
  • 19Software Architecture for Intelligent Systems
  • 20Scientific Software Documentation
  • 21Research Transfer: Intelligent-System Software
  • 22Industry Practice & Engineering Decision Record
  • 23Module Principle & Transfer

Practical implementation

  • 01Laboratory 01 - Scientific Codebase from Scratch
  • 02Laboratory 02 - Memory & Vectorisation Investigation
  • 03Laboratory 03 - Debugging & Failure Injection
  • 04Laboratory 04 - Reproducible Experiment Pipeline

Tools and technologies

  • Linux VM/container
  • Python + C/C++
  • profiling
  • testing
  • reproducible environments

Project

Laboratory 04 - Reproducible Experiment Pipeline

Expected learning outcomes

  • 01Trace a simple program from source-level idea Ann to execution and state changes.
  • 02Reason about types, representation and memory.
  • 03Design modular scientific software.
  • 04Write tests as executable specifications.
  • 05Use debugging and failure isolation systematically.
  • 06Measure software performance rather than guessing.
  • 07Understand vectorisation and locality.
  • 08Build reproducible environments and workflows. Ve
  • 09Engineer software suitable for later intelligent- system modules.

Assessment

  • 01Daily practice — 20%
  • 02Labs — 30%
  • 03Research reproduction — 10%
  • 04Integrated project — 20%
  • 05Technical viva — 15%
  • 06Engineering record — 5%

Source document: TUTOR Frontier Module 05 Scientific Programming Software Engineering for Intelligent Systems