Deep Learning & Neural Networks

Master deep neural architectures, exact gradient backpropagation calculus, and production PyTorch systems.

This course provides mathematical foundations and high-throughput engineering paradigms for deep neural networks. Students derive backward gradient equations across non-linear manifolds and implement distributed optimization algorithms. Advanced topics include architectural design of Transformers and deployment optimization with TensorRT and ONNX Runtime.

90h + 48h 2 certificates available

Deep LearningPyTorchBackpropagation CalculusTransformer ArchitecturesDistributed Training

Curriculum

2 complete tracks. Study any or all — each has its own exam and certificate.

A mathematically rigorous exploration of deep learning theory, covering multivariable calculus of backpropagation, empirical risk minimization, representation theory, and modern attention mechanics.

Formalize feedforward neural representations, auto-differentiation topologies, and non-con

Reverse-Mode Automatic Differentiation & Jacobian Calculus Lab40 min
Stochastic Gradient Dynamics & Hessian Curvature Analysis Lab45 min
Generalization Bounds & Spectral Norm Regularization Lab35 min

Careers this prepares you for

Deep Learning Research Engineer
Computer Vision Specialist
NLP Infrastructure Engineer
Machine Learning Systems Architect

How you study

Every lesson pairs an in-depth video with written notes, key takeaways, a hands-on lab, exercises and a quiz. Work at your own pace, then sit the final examination for your Erudex certificate.

Deep Learning & Neural Networks

Curriculum

Module 1Foundations of Optimization & Computational Graphs

Module 2Equivariance, Convolutions, and Sequence Representation

Module 3Mathematical Foundations of Self-Attention & Latent Diffusion

Module 4Approximation Theory, Representation Capacity, and Generalization

Module 5Loss Landscape Geometry and Advanced Training Dynamics

Module 6Graph Learning and Geometric Representation

Module 7Distributed Training and High-Throughput Systems

Module 8Evaluation, Uncertainty, and Robustness

Module 9Model Compression and Production Inference

Module 10Capstone: Reproducible Adaptation and Deployment of a Deep Learning System

Careers this prepares you for

How you study