Understanding Deep Neural Networks Training Course

Part 1 – Deep Learning and DNN Concepts

Introduction to AI, Machine Learning & Deep Learning

• The history, core concepts, and common applications of artificial intelligence, moving beyond theoretical abstractions.
• Collective Intelligence: aggregating knowledge shared among numerous virtual agents.
• Genetic algorithms: evolving populations of virtual agents through selection processes.
• Standard Machine Learning: defining key concepts.
• Task categories: supervised learning, unsupervised learning, and reinforcement learning.
• Action types: classification, regression, clustering, density estimation, and dimensionality reduction.
• Examples of Machine Learning algorithms: Linear regression, Naive Bayes, and Random Forests.
• Machine Learning vs. Deep Learning: identifying problems where Machine Learning remains the current state of the art (e.g., Random Forests & XGBoosts).

Basic Concepts of a Neural Network (Application: multi-layer perceptron)

• Review of essential mathematical foundations.
• Definition of a neural network: classical architecture, activation functions, and input weighting.
• The concept of network depth.
• Defining network learning: cost functions, back-propagation, Stochastic gradient descent, and maximum likelihood.
• Modeling neural networks: adapting input and output data structures to problem types (regression, classification, etc.) and addressing the curse of dimensionality.
• Distinguishing between multi-feature data and signals; selecting appropriate cost functions based on data characteristics.
• Function approximation using neural networks: theoretical presentation and case examples.
• Distribution approximation using neural networks: theoretical presentation and case examples.
• Data Augmentation: techniques for balancing datasets.
• Generalizing the results obtained from neural networks.
• Initialization and regularization of neural networks: L1 / L2 regularization and Batch Normalization.
• Optimization and convergence algorithms.

Standard ML / DL Tools

This section provides an overview of tools, highlighting their advantages, disadvantages, position in the ecosystem, and usage.

• Data management tools: Apache Spark, Apache Hadoop Tools.
• Machine Learning libraries: NumPy, SciPy, Scikit-learn.
• High-level DL frameworks: PyTorch, Keras, Lasagne.
• Low-level DL frameworks: Theano, Torch, Caffe, TensorFlow.

Convolutional Neural Networks (CNN).

• Overview of CNNs: fundamental principles and primary applications.
• Core CNN operations: convolutional layers and the use of kernels.
• Padding and stride, feature map generation, and pooling layers. Extensions to 1D, 2D, and 3D contexts.
• Examination of various CNN architectures that have achieved the state of the art in classification tasks.
• Image processing: LeNet, VGG Networks, Network in Network, Inception, and ResNet. Analysis of innovations introduced by each architecture and their broader applications (e.g., 1x1 convolutions or residual connections).
• Application of attention models.
• Practical application to standard classification cases (text or image).
• CNNs for generation: super-resolution and pixel-to-pixel segmentation.
• Key strategies for enhancing feature maps in image generation.

Recurrent Neural Networks (RNN).

• Overview of RNNs: fundamental principles and typical applications.
• Basic RNN mechanics: hidden state activation, back propagation through time, and the unfolded network representation.
• Evolution towards Gated Recurrent Units (GRUs) and LSTM (Long Short-Term Memory).
• Analysis of various states and the improvements brought by these advanced architectures.
• Addressing convergence issues and the vanishing gradient problem.
• Classic architectures: Time series prediction, classification, and other sequential tasks.
• RNN Encoder-Decoder architectures and the integration of attention models.
• NLP applications: word/character encoding and machine translation.
• Video Applications: predicting the next frame in a video sequence.

Generative models: Variational AutoEncoder (VAE) and Generative Adversarial Networks (GAN).

• Introduction to generative models and their relationship with CNNs.
• Auto-encoders: dimensionality reduction and limited scope generation.
• Variational Auto-encoders: as a generative model approximating data distribution. Definition and utilization of latent space, the reparameterization trick, and observed applications and limitations.
• Generative Adversarial Networks: core fundamentals.
• Dual Network Architecture (Generator and Discriminator) with alternating learning strategies and available cost functions.
• GAN convergence and common challenges encountered.
• Enhanced convergence techniques: Wasserstein GAN, BEGAN, and Earth Moving Distance.
• Applications in image and photograph generation, text creation, and super-resolution.

Deep Reinforcement Learning.

• Introduction to reinforcement learning: controlling an agent within a defined environment.
• Defining states and possible actions.
• Utilizing neural networks to approximate the state function.
• Deep Q Learning: experience replay and its application to video game control.
• Policy optimization: on-policy vs. off-policy methods, Actor-critic architecture, and A3C.
• Applications: controlling single video games or digital systems.

Part 2 – Theano for Deep Learning

Theano Basics

• Introduction.
• Installation and configuration.

Theano Functions

• Handling inputs, outputs, updates, and givens.

Training and Optimizing Neural Networks with Theano

• Neural Network Modeling.
• Logistic Regression.
• Implementing Hidden Layers.
• Training the network.
• Computing predictions and Classification.
• Optimization techniques.
• Log Loss calculation.

Model Testing

Part 3 – DNN using TensorFlow

TensorFlow Basics

• Creating, initializing, saving, and restoring TensorFlow variables.
• Feeding, reading, and preloading TensorFlow Data.
• Leveraging TensorFlow infrastructure to train models at scale.
• Visualizing and evaluating models using TensorBoard.

TensorFlow Mechanics

• Data Preparation.
• Downloading data.
• Managing Inputs and Placeholders.
• Building Graphs.
  • Inference.
  • Loss calculation.
  • Training process.
• Training the Model.
  • The Graph structure.
  • The Session context.
  • The Train Loop.
• Evaluating the Model.
  • Building the Eval Graph.
  • Interpreting Eval Output.

The Perceptron

• Activation functions.
• The perceptron learning algorithm.
• Binary classification using the perceptron.
• Document classification with the perceptron.
• Limitations of the perceptron model.

From Perceptrons to Support Vector Machines

• Kernels and the kernel trick.
• Maximum margin classification and support vectors.

Artificial Neural Networks

• Nonlinear decision boundaries.
• Feedforward and feedback artificial neural networks.
• Multilayer perceptrons.
• Minimizing the cost function.
• Forward propagation.
• Back propagation.
• Strategies to improve neural network learning.

Convolutional Neural Networks

• Objectives.
• Model Architecture.
• Core Principles.
• Code Organization.
• Launching and Training the Model.
• Evaluating a Model.

Brief introductions to the following modules will be provided, subject to time availability:

TensorFlow – Advanced Usage

• Threading and Queues.
• Distributed TensorFlow.
• Writing documentation and sharing models.
• Customizing Data Readers.
• Manipulating TensorFlow Model Files.

TensorFlow Serving

• Introduction.
• Basic Serving Tutorial.
• Advanced Serving Tutorial.
• Serving Inception Model Tutorial.