Neural network programme catalogue

NeuralCourseHub maintains a curated course catalogue designed as progressive learning pathways. Each programme carries a reference code (NCH-001 through NCH-006), documented module hours, and a capstone exercise aligned with vocational training standards in Ontario. Enrolment is open to Canadian residents and international learners who can participate in scheduled live online sessions (times listed in Eastern Time unless noted).

NCH-001

Neural Network Fundamentals

Your entry point into artificial neural networks. This programme covers perceptrons, multilayer architectures, activation functions, and forward propagation with hands-on PyTorch exercises. Learners implement a basic classifier, explore weight initialisation, and interpret tensor shapes through guided module work. You will understand how layers stack to form expressive models and why non-linear activations matter — foundational knowledge for every subsequent pathway tier. Suitable for developers and analysts with working Python skills who have not yet trained neural models in code.

Duration: 8 weeks Format: Live online cohort Prerequisite: Python basics

Indicative fee: $1,450 – $1,650 CAD

Course enrolment inquiry

NCH-002

Backpropagation & Optimisation Deep Dive

Move from forward passes to the mechanics of learning. This programme dissects backpropagation, gradient computation, vanishing gradient challenges, and optimiser families including SGD, momentum, and Adam. Learners experiment with learning rate schedules, batch size effects, and regularisation techniques such as dropout and weight decay. Module exercises include training curves analysis and debugging common convergence failures. Completion prepares you for architecture-specific programmes with confidence in the training loop itself.

Duration: 6 weeks Format: Live online cohort Prerequisite: NCH-001 or equivalent

Indicative fee: $1,350 – $1,550 CAD

Course enrolment inquiry

NCH-003

Convolutional Neural Networks (CNN) Course

Specialise in image-oriented deep learning. Study convolution operations, filter banks, pooling layers, and classic architectures adapted for educational scope. Implement CNN classifiers on curated datasets, interpret feature maps at a conceptual level, and practice transfer learning introductions. The capstone asks learners to document architecture choices, training metrics, and limitations — emphasising evaluation honesty over leaderboard chasing. Hybrid sessions may be available at our Waterloo training room by appointment.

Duration: 10 weeks Format: Hybrid (online + optional in-person) Prerequisite: NCH-002 recommended

Indicative fee: $1,750 – $2,050 CAD

Course enrolment inquiry

NCH-004

Recurrent Neural Networks & Sequence Models

Address sequential data with RNN fundamentals, LSTM cells, and practical text classification workflows. Learners build sequence models in PyTorch, handle padding and masking, and evaluate performance on time-series and language tasks at introductory depth. The programme contrasts recurrent approaches with emerging transformer methods so you understand when each family applies. Module readings include ethical considerations for language model applications without venturing into unsupervised large-scale training claims beyond educational scope.

Duration: 8 weeks Format: Live online cohort Prerequisite: NCH-002 recommended

Indicative fee: $1,650 – $1,900 CAD

Course enrolment inquiry

NCH-005

Transformer Architecture Fundamentals

Explore the architecture that reshaped modern NLP and increasingly vision tasks. Study self-attention, positional encoding, encoder-decoder structures, and the intuition behind transformer blocks — taught at fundamentals depth appropriate for vocational learners, not research fellowship level. Exercises include implementing simplified attention components and fine-tuning introductory pretrained models in a supervised educational context. Learners leave with vocabulary to read technical papers and participate in engineering discussions responsibly.

Duration: 10 weeks Format: Live online cohort Prerequisite: NCH-003 or NCH-004

Indicative fee: $1,850 – $2,150 CAD

Course enrolment inquiry

NCH-006

Neural Model Evaluation & Deployment Overview

Close your pathway with rigorous evaluation and a pragmatic introduction to deployment concepts. Topics include metric selection, confusion matrices, calibration awareness, bias-variance discussion, model serialization basics, and inference serving overview — without claiming production DevOps mastery in a single programme. The capstone integrates a complete model lifecycle document suitable for portfolio presentation. Ideal for learners who have completed at least two architecture programmes and want structured closure before independent project work.

Duration: 6 weeks Format: Live online cohort Prerequisite: Two prior NCH programmes

Indicative fee: $1,400 – $1,600 CAD

Course enrolment inquiry