Diploma in Classical and Deep RL for Dynamic Systems
Online postgraduate diploma of 800 hours in classical and deep reinforcement learning for dynamic systems. 8 modules covering Markov decision processes, tabular methods, function approximation, value-based deep RL, policy gradient and actor-critic, model-based RL and sim-to-real transfer in line with ISO/IEC 22989:2022, ISO/IEC 23053:2022 and IEC 61508:2010, with an assigned technical tutor.
Diploma in Classical and Deep RL for Dynamic Systems — studying from New Zealand
New Zealand's data and artificial intelligence sector is growing, with strong activity in Auckland, Wellington and Christchurch across fintech, agritech, health analytics and government data services. Universities including the University of Auckland, the University of Otago and Victoria University of Wellington run leading research in machine learning and computational modelling. Simulation and analytics skills are widely applied in agriculture, energy and environmental modelling, and employers include government agencies such as Stats NZ as well as technology firms and consultancies.
Fees for New Zealand: NZ$2,550 (New Zealand dollar), with instalments and employer sponsorship available.
Relevant bodies in New Zealand
Typical roles
Questions from New Zealand
Are data and AI engineering roles regulated in New Zealand, and how does that affect a private qualification?
Data and AI are not regulated professional titles in New Zealand, though privacy obligations fall under the Office of the Privacy Commissioner. Engineers who want chartered status can pursue registration through Engineering New Zealand. SEIUM's qualifications are private, not state degrees, and not accredited by a national body, so employers decide how much they value them.
Which New Zealand employers hire data, AI and simulation engineers?
Demand is found in banks and insurers, government agencies, healthcare, energy and utilities, technology firms, and consultancies. Employers usually assess portfolios, practical projects and technical interviews more than credentials. A private online programme can help build relevant skills, but it does not guarantee a position or a specific salary.
Can I complete this online from New Zealand in English while working?
Yes. The programmes are fully online and taught in English, so you can study from New Zealand without a student visa. You will have an assigned technical tutor and complete a final project. Specialist courses are 300 hours, diplomas 800 and master's 1,900. Fees can be paid in instalments, and some employers sponsor study, so ask yours early.
Take the controls.
Run a real engineering trade-off from this field. Move the sliders, watch the model react in real time and beat the challenge.
What you will be able to do
Model dynamic systems as Markov decision processes.
Implement tabular temporal-difference and Monte Carlo methods.
Apply function approximation and deep networks to value learning.
Train actor-critic agents in continuous action spaces.
Integrate learned dynamic models into planning and control.
Design simulation-to-real transfer strategies.
Evaluate agents with reproducible metrics and safety criteria.
Who it's for
- Reinforcement learning engineers.
- Intelligent control engineers looking beyond classical PID and MPC.
- Robotics researchers.
- Data scientists working in optimisation.
- Autonomous systems engineers.
- Simulation developers.
- Engineering or science graduates seeking to enter industrial robotics.
What you'll work with
- ISO/IEC 22989:2022 (AI concepts and terminology)
- ISO/IEC 23053:2022 (Framework for AI systems using machine learning)
- IEC 61508:2010 (Functional safety of E/E/PE safety-related systems)
- ISO 26262:2018 (Road vehicles: functional safety)
- Regulation (EU) 2024/1689 (EU Artificial Intelligence Act)
- ISO/IEC 23894:2023 (AI risk management)
- IEC 61131-3:2013 (PLC programming languages)
- Gymnasium
- Stable-Baselines3
- MuJoCo
- Ray RLlib
- NVIDIA Isaac Lab
8 modules
1.1States, actions, rewards and transitions1.2Discounted return and horizon1.3Bellman equations for value and action-value1.4Value iteration and policy iteration1.5Dynamic programming and contraction1.6Partially observable MDPs and belief states1.7Discretisation of continuous dynamic systems1.8Reward function design1.9Relationship with LQR optimal control1.10Workshop: inverted pendulum solved by dynamic programming
2.1Monte Carlo policy evaluation2.2Temporal differences TD(0) and TD(lambda)2.3Eligibility traces2.4SARSA and on-policy control2.5Q-learning and off-policy control2.6Epsilon-greedy and softmax exploration2.7Double Q-learning and maximisation bias2.8Dyna-Q and model-based planning2.9Convergence and Robbins-Monro conditions2.10Lab: Q-learning on discretised MountainCar
3.1Linear approximation and basis functions3.2Tile coding and Fourier bases3.3Semi-gradient TD with linear approximation3.4The deadly triad of instability3.5Least-squares TD and LSPI3.6Neural networks as approximators3.7Normalisation of observations and rewards3.8Generalisation across states3.9Hyperparameter tuning and seeds3.10Practical exercise: CartPole control with linear approximation
4.1Deep Q-Network and experience replay4.2Target networks and stabilisation4.3Double DQN and Dueling DQN4.4Prioritised experience replay4.5Rainbow and combining improvements4.6Distributional RL: C51 and QR-DQN4.7Exploration with parametric noise4.8Discrete actions in physical systems4.9Diagnosing learning curves4.10Lab: DQN with Stable-Baselines3
5.1Policy gradient theorem and REINFORCE5.2Baselines and variance reduction5.3Actor-critic A2C and A3C5.4Generalised advantage estimation (GAE)5.5TRPO and trust regions5.6PPO with ratio clipping5.7DDPG for continuous actions5.8TD3 and critic overestimation5.9Soft Actor-Critic and maximum entropy5.10Lab: SAC on a robotic arm in MuJoCo
6.1Learning dynamic models with neural networks6.2PETS probabilistic ensembles6.3Model predictive control with a learned model6.4MBPO and short rollouts6.5World models and Dreamer6.6System identification versus learned models6.7Compounding error and planning horizon6.8Sample efficiency in real systems6.9Combining physics and data6.10Practical exercise: quadrotor control with a learned model
7.1The reality gap in dynamics and sensors7.2Domain randomisation of physical parameters7.3Parameter identification and calibration7.4Offline RL with CQL and IQL7.5Safety when deploying on hardware7.6Delays, noise and actuator saturation7.7Gymnasium and Isaac Lab environments7.8Evaluation and reproducible benchmarks7.9AI regulation and functional safety7.10Case study: temperature control of a thermal process
8.1System selection and MDP formulation8.2Building the simulation environment8.3Reward and constraint design8.4Training baseline agents8.5Comparison of algorithms and seeds8.6Robustness analysis against disturbances8.7Hardware transfer plan8.8Reproducible documentation of the experiment8.9Report and presentation8.10Final project: final review with the tutor
How you'll learn
100% online training on the SEIUM campus, with downloadable material, recorded classes and an assigned technical tutor. There are no fixed timetables: you progress at your own pace within each intake, so the programme is compatible with work.
Each module combines theory, guided reading of the original standard and applied exercises. The working software is Gymnasium, Stable-Baselines3, MuJoCo, Ray RLlib and NVIDIA Isaac Lab, in free, academic or evaluation versions. Practical work is based on real procedures for control of pendulums and underactuated systems in simulation, tuning of thermal process controllers and simulated robotic locomotion and manipulation. Assessment is continuous by module plus the final project.
Standards covered: ISO/IEC 22989:2022, ISO/IEC 23053:2022, IEC 61508:2010, ISO 26262:2018, Regulation (EU) 2024/1689, ISO/IEC 23894:2023, IEC 61131-3:2013.
Build it end to end
Your final project follows the real engineering workflow, step by step, reviewed by your tutor.
Where this takes you
Requirements
- Prior technical training or professional experience in the sector. No university degree is required.
- Internet connection and a computer capable of running the programme software.
- English sufficient to read technical standards.
Ready to start?
Send your application today. Our admissions team replies within 72 hours with your study plan and funding options.
Frequently asked questions
The programme totals 800 hours over 6–8 months. You set your own pace within each intake, and your tutor helps you plan the workload around your job.
You work with Gymnasium, Stable-Baselines3, MuJoCo, Ray RLlib and NVIDIA Isaac Lab, using free, academic or evaluation versions. Your tutor indicates the free alternative wherever one exists.
A SEIUM Institute diploma, awarded on passing the continuous assessment and the final project. It is SEIUM's own qualification, not an official state degree; its recognition depends on the programme completed and the regulations of each country.
Yes. Modules passed are recognised in a higher programme in the same area (postgraduate diploma or master's). Ask about the recognition policy when you enrol.
Prior technical training or professional experience in the sector (no university degree required), a computer able to run the programme software and English sufficient to read technical standards.
Start dates are set by SEIUM's faculty for each intake. Our admissions team will confirm the next available date when you apply.
Related programmes
Diploma in Uplift Modeling and Causality-Based Decision Making
Online postgraduate diploma of 800 hours in uplift modelling and causality-based decision-making. 8 modules covering potential outcomes and causal graphs, experimental design, meta-learners, causal trees and…
Diploma in Spectrum, Filters and Detection of Rare Events
Online postgraduate diploma of 800 hours in spectrum, filters and rare event detection. 8 modules covering sampling, spectral analysis, digital filter design, classical and machine learning…
Diploma in Roadmap, Data SLAs and Service Contracts
Online postgraduate diploma of 800 hours in data roadmaps, data SLAs and service contracts, with 8 modules covering data product management, data roadmaps, data service levels,…
More in Data, AI, Simulation & Analytics
All Data, AI, Simulation & Analytics programmes → · How to apply · Fees & financing · Scholarships