Advanced Deep Learning Professional (ADLP)

Course Information

  • Duration: 5 Day / 40 Hours
  • Certification: Participants will receive a Certificate of Competency upon successfully completing the course and passing the examination
  • Who Should Attend: Data Scientist, Data Analyst, Analyst, System Analyst, Technologist, System Engineer, IT Professionals and Anyone seeking to acquire advanced knowledge on Deep Learning

Course Objective

Advanced Deep Learning Professional (ADLP) is designed for anyone interested in acquiring the advanced knowledge and skills required to implement / manage Deep Learning an organization

Pre-Requisite

It is preferred that participants have some prior experience in Python Programming, Data Analytics or successfully completed and received a Certificate of Competency in Python Programming Essentials (PPE).

Examination

Participants are required to attempt an examination upon completion of course. This exam tests a candidate’s knowledge and skills related to Deep Learning based on the syllabus covered

Module 1 INTRODUCTION TO DEEP LEARNING

  • What is Artificial Intelligence and Machine Learning?
  • Understanding Learning Representation of Data
  • Fundamentals of Deep Learning
  • How Deep Learning Works
  • Deep Learning and Its Application
  • Future of Deep Learning

  • Module 2 MACHINE LEARNING ESSENTIALS

  • Probabilistic Modeling
  • Kernel Methods
  • Decision trees, Random Forests, and Gradient Boosting Machines
  • Understanding Neural Networks
  • What Makes Deep Learning Different
  • Hardware, Data, and Algorithms of Deep Learning

  • Module 3 Data Representation in Neural Networks

  • Scalars, Vectors, Matrices
  • 3D Tensors and higher-dimensional tensors
  • Key Attributes
  • Manipulating Tensors in Numpy
  • The Notion of Data Batch
  • Examples of Data Tensors
  • Vector Data
  • Timeseries Data or Sequence Data
  • Image and Video Data

  • Module 4 Neural Networks: Tensor Operations

  • Element-wise Operations
  • Broadcasting
  • Tensor Dot
  • Tensor Reshaping
  • Geometric Interpretation of Tensor Operations
  • Geometric Interpretation of Deep Learning

  • Module 5 Neural Networks: Gradient-Based Operations

  • Introduction to Derivative
  • Derivative of a Tensor Operation: Gradient
  • Stochastic Gradient Descent
  • The Backpropagation Algorithm

  • Module 6 Structure of a Neural Network

  • What are the Layers for Deep Learning?
  • Neural Network Models
  • Core Elements to Configuring a Learning Process
  • Introduction to Keras, TensorFlow, Theano, and CNTK
  • Brief Overview of Keras

  • Module 7 Getting Ready for Deep Learning

  • Key Considerations
  • Setting up Jupyter Notebooks
  • Setting up of Keras
  • Deep Learning in a Cloud
  • Identifying the Best GPU for Deep Learning

  • Module 8 Binary Classification

  • Preparing the Data
  • Building Network
  • Validating Approach
  • Using a Trained Network to Generate Predictions on New Data

  • Module 9 Multi-Class Classification

  • Preparing the Data
  • Building Network
  • Validating Approach
  • Generate Predictions on New Data
  • Alternative Ways to Handle Labels and Loss
  • Importance of Having Sufficiently Large Intermediate Layers

  • Module 10 Regression Model

  • Preparing the Data
  • Building Network
  • Validating Approach using K-Fold Validation

  • Module 11 Deep Learning for Computer Vision

  • Introduction to Convnets
  • Understanding Convolution Operation and Max Pooling Operation
  • Training a Convnet on a Small Dataset
  • Understanding the Relevance of Deep Learning for Small-Data Problems
  • Downloading the Data and Building the Network
  • Data Pre-processing and Data Augmentation
  • Visualizing Intermediate Activations
  • Visualizing Convnet Filters
  • Visualizing Heatmaps of Class Activation

  • Module 12 Deep Learning for Text and Sequences

  • Encoding of Words or Characters
  • How is Word Embedding Being Used?
  • From Raw Text to Word Embeddings
  • Understanding Recurrent Neural Networks
  • What is LSTM and GRU Layers?
  • What is a First Recurrent Baseline
  • Using Recurrent Dropout to Fight Overfitting
  • Stacking Recurrent Layers
  • Using Bidirectional RNNs
  • Understanding 1D Convolution for Sequence Data
  • 1D Pooling for Sequence Data
  • Implementing a 1D Convnet
  • Combing CNNs and RNNs to Process Long Sequences

  • Module 13 Advanced Deep Learning Techniques

  • Introduction to Functional APIs
  • Multi-Input and Multi-Output Modes
  • Directed Acyclic Graphs of Layers
  • Layer Weight Sharing
  • Using Keras Callbacks and TensorBoard
  • Understanding TensorFlow Visualization Framework
  • Advanced Architecture Patterns
  • Hyperparameter Optimization
  • Model Ensembling

  • Module 14 Generative Deep Learning

  • Text Generation with LSTM
  • Implementing Deep Dream in Keras
  • Neural Style Transfer in Keras
  • Generating Images with Variational Autoencoders
  • What is a Generative Adversarial Networks
  • Training DCGAN