Deep Learning Course Structure

About This Course

4.9 3.2K ratings

This Course Is Designed To Give An Extensive Knowledge To The Learner In The Field Of Deep Learning. The Key Emphasis Of The Course Is To Enable The Learner Understand Deep Convolutional Neural Networks, Understand The Regularization And Optimization Techniques To Train A CNN And Attain Hands On Experience In Computer Vision.

Although The Course Starts With Python, But The Main Emphasis IS Problem Solving And Aglorithmic Design In Python.We Try To Give A Brief Introduction To Competitive Programming In First Week.This Style Of Teaching A Programming Language Is Vey Effective Language Is Very Effective And Provides An Accelerated Learning To The Learner. Remaininng Course Is Divided Into Tutorials Where We Try To Provide Hand On Experience In Numpy , Pandas , Sikit Learn , Keras And Statsmodal , While On Theoritical Side , The Course Will Focus On Fundamental Mathematical Constructs Like Linear Algebra . Probability And Information Theory To Build A Solid Backround For Deep learning.

In The Macine Learning Part , We Aim To Explain In Great Theoritical Details Of Some Machine Learning Algorithmic Which Are Critical For Understanding Of Deep Neural Networks And Other Fundamentals Necessary For Machine And On Practical Side , We Focus To Use Scikit Learn To Quickly Work On Many Other Machine Learning Algorithms. The Course Then MOves To Deep Forward Neural Networks , Backpropogation To Train Them And Other Important Topics Like Regularization And Optimization And Finally To Convolutional Neural Networks Where On Practical Side We Will Focus On Diffrent Computer Vision Use Cases.

Syllabus - What you will learn from this course

WEEK
1

Competitive Programming With Python

4 Lectures 8 Tutorials , 30 Hours

THEORY


  • INTRODUCTION TO ALGORITHMIC COMPLEXITY

  • PYTHON SYNTAX

  • CONTROL FLOW BASICS AND PRACTICE QUESTIONS



  • TUTORIALS


  • BACKTRACKING

  • DYNAMIC PROGRAMMING

  • CLASSICAL ARTIFICIAL INTELLEGENCE PROBLEMS


  • WEEK
    2

    Linear Algebra

    3 Lectures 3 Tutorials , 12 Hours

    THEORY


  • SCALARS, VECTORS, MATRICES AND TENSORS

  • LINEAR TRANSFORMATIONS, BASIS, SPAN, NORMS, MATRIX INVERSION AND DETERMINANTS

  • SPECIAL KIND OF MATRICES

  • EIGEN DECOMPOSITION AND SINGULAR VALUE DECOMPOSITION

  • THE TRACE OPERATOR

  • PRINCIPAL COMPONENTS ANALYSIS



  • TUTORIALS


  • MATRIX MANIPULATION WITH NUMPY


  • WEEK
    3

    Probability and Information Theory

    8 Lectures 4 Tutorials 30 Hours (4 Extra Classes)

    THEORY


  • MEASURES OF CENTRALITY AND DISPERSIONS

  • SET THEORY AND P&C

  • JOINT, MARGINAL AND CONDITIONAL PROBABILITY

  • MUTUAL EXCLUSION, INDEPENDENCE AND CONDITIONAL INDEPENDENCE

  • BAYSE THEROEM

  • EXCEPTION VARIANCE AND COVARIANCE

  • DISCRETE PROBABILITY DISTRIBUTIONS

  • CENTRAL LIMIT THEOREM AND CONTINUOUS PROBABILITY DISTRIBUTION

  • INFORMATION THEORY

  • SIGMOID, LOGIT & PROBIT, CROSS ENTROPY AND SOFTMAX

  • CONVOLUTION

  • STRUCTURED PROBABILISTIC MODELS



  • TUTORIALS


  • REVISION OF P&C

  • PROBABILITY PRACTICE QUESTIONS

  • STATSMODEL


  • WEEK
    4

    Machine Learning

    10 Lectures 14 Tutorials 60 Hours

    THEORY


  • LEARNING ALGORITHMS

  • CAPACITY, OVERFITTING AND UNDERFITTING

  • HYPERPARAMETER TUNING AND VALIDATION SETS

  • ESTIMATORS, BIAS AND VARIANCE

  • MAXIMUM LIKELIHOOD ESTIMATION

  • GRADIENT BASED OPTIMIZATION AND STOCASTIC GRADIENT DESCENT

  • DERIVING MULTIVARIATE BLINEAR REGRESSION

  • DERIVING LOGISTIC REGRESSION AND ALPHA PERCEPTRONES FOR CLASSIFICATION

  • DERIVNG SUPPORT VECTOR MACHINES AND KERNAL TRICK FOR NON-LINEAR LEARNING

  • CHALLENGES MOTIVATING DEEP LEARNING



  • TUTORIALS


  • PANDAS AND SCIKIT LEARN

  • LINEAR REGRESSION

  • LOGISTIC REGRESSION

  • SUPPORT VECTOR MACHINES

  • DECISION TREES AND GRADIENT BOOSTING TREES

  • AGLOMERATIVE AND K MEANS CLUSTERING


  • WEEK
    5

    Feed Forward Neural Network

    2 Lectures 4 Tutorials 14 Hours , 3 quizzes

    THEORY


  • LEARNING XOR

  • KEY-NOTES TO ARCHITECTURE DESIGN OF NEURAL NETWORKS

  • BACKPROPAGATION

  • TUTORIALS


  • INTRODUCTION TO KERAS


  • WEEK
    6

    REGULARIZATION FOR DEEP LEARNING

    6 Lectures 0 Tutorials , 14 Hours , 3 quizzes

    THEORY


  • Parameter Norm Penalties

  • Norm Penalties as Constrained Optimization

  • Regularization and Under-Constrained Problems

  • Dataset Augmentation

  • Noise Robustness

  • Semi-Supervised Learning

  • Multitask Learning

  • Early Stopping

  • Parameter Tying and Parameter Sharing

  • Sparse Representations

  • Bagging and Other Ensemble Methods

  • Dropout

  • Adversarial Training

  • Tangent Distance, Tangent Prop and ManifoldTangent Classifier


  • WEEK
    7

    Optimization for Training Deep Models

    4 Lectures 0 Tutorials 10 Hours , 3 quizzes

    THEORY


  • How Learning Differs from Pure Optimization

  • Challenges in Neural Network Optimization

  • Basic Algorithms

  • Parameter Initialization Strategies

  • Algorithms with Adaptive Learning Rates

  • Approximate Second-Order Methods

  • Optimization Strategies and Meta-Algorithms


  • WEEK
    8

    Convolutional Networks

    5 lectures 10 tutorials 40 Hours , 3 quizzes

    THEORY


  • The Convolution Operation

  • Motivation

  • Pooling

  • Convolution and Pooling as an Infinitely Strong Prior

  • Variants of the Basic Convolution Function

  • Structured Outputs

  • Data Types

  • Efficient Convolution Algorithms

  • Random or Unsupervised Features

  • The Neuroscientific Basis for Convolutional Networks

  • Convolutional Networks and the History of Deep Learning


  • 4.8

    12 Reviews

    Top Reviews


    Dec 11 -2018
    BY ME

    A career changer course, thanks to the hand-one which is second to none, I have gained experience which on other online course can produce, thanks to IBM for this course which timely and excellent.

    Dec 11 -2018
    BY ME

    A Career Changer course, thanks the hand-ons which is second to none, I have gained experience which on other online course can produce, thanks to IBM for this course which timely and excellent.

    Instructors