Certified Training   

 

AI & Deep Learning  Training


Program duration:   

Non-IT professionals : 10 months (546h)

IT professionals : 7 months (364h)

Objectives:

 Gain expertise in Frameworks such as TensorFlow, PyTorch, and state-of-the-art AI architectures. 

 Apply Deep Learning in real-world applications such as Computer Vision, NLP, Generative AI, and Reinforcement Learning.  

Modules:

Fundamentals of Deep Learning (28h)/(42h)   

  • Introduction to Python programming
  • Introduction to Deep Learning & Neural Networks
    • Supervised vs. Unsupervised Learning
    • Introduction to Perceptrons, Activation Functions
  • Mathematical Foundations
    • Linear Algebra
    • Probability
    • Calculus for Deep Learning
    • Backpropagation
    • Optimization (SGD, Adam, RMSProp)

✅  Deep Neural Networks & Optimization  ​ (28h)/(42h)

  • Building Deeper Architectures
    • Batch Normalization
    • Dropout
    • Regularization Techniques
  • Hyperparameter Tuning & Model Optimization
    • Learning Rate Scheduling,
    • Weight Initialization
    • Optimization Tools
    • Weights & Biases
  • Capstone project: DNN Application (28h)/(42h)

Convolutional Neural Networks (CNNs) & Computer Vision (28h)/(42h)

  • Convolutional Layers, 
  • Pooling
  • Feature Extraction
  • Advanced Architectures (ResNet, EfficientNet, Vision Transformers)
  • Image Processing & Object Detection
    • YOLO
    • Faster R-CNN, 
    • Semantic Segmentation (U-Net)
  • Capstone Project: Computer Vision Application (28h)/(42h)
    • Facial Recognition, Medical Image Analysis, Autonomous Driving (28h)/(42h)

 

Natural Language Processing (NLP) & Transformers (28h)/(42h)

  • Introduction to NLP & RNNs
  • Sequence Models: LSTMs, GRUs, Attention Mechanism
  • NLP Models (Transformer Architecture: BERT, GPT, T5, LLaMA)
  • Text Generation (Fine-tuning Pretrained Models)
  • Capstone project: NLP Application (28h)/(42h)

✅  Generative AI & Advanced Deep Learning ​ (28h)/(42h)

  • Generative Adversarial Networks (GANs)
  • GANs and Diffusion Models
  • Text-to-Image Generation
  • Image-to-Image Generation
  • Capstone project: Generative AI Application (28h)/(42h)

✅  Deep Reinforcement Learning (DRL) (28h)/(42h)

  • Introduction to Deep Q-Networks (DQN)
  • Experience Replay
  • Target Network
  • Deep Q-Learning, Policy Gradient Methods
  • Capstone project: DRL in Real-World Applications (28h)/(42h)
  • DRL in robotics, DRL in finance and trading, DRL in recommendation systems

Reinforcement Learning & AI Agents  (28h)/(42h)  

  • Components of RL: Agent, Environment, States, Actions, Rewards
  • Differences between RL, Supervised, and Unsupervised Learning
  • Model-Free Learning: Temporal Difference (TD) Learning
    • TD(0) Algorithm
    • SARSA (State-Action-Reward-State-Action)
    • Q-Learning
    • Difference between On-Policy and Off-Policy Learning


  • Capstone project: RL in Real-World Applications (28h)/(42h)

​RL in robotics; RL in finance and trading; RL in recommendation systems

NVIDIA Certifications ​


 Fundamentals of Deep Learning

 Building Transformer-Based Natural Language Processing Applications 

 Building LLM Applications With Prompt Engineering

 Building RAG Agents with LLMs

 ​

  Efficient LLM Customization

 Rapid Application Development with LLMs

 Building Conversational AI Application

 Computer Vision for Industrial Inspection

 Generative AI with Diffusion Models

 (+216) 98 106 016  -(+216) 98 270 400

   training-center@horizon-tech.tn