Deep Neural Networks using Python Online Course
TensorFlow, Google’s open-source deep learning library, is one of the most widely used tools in artificial intelligence and machine learning today. Mastering it is essential for anyone pursuing deep learning. In this course, you will learn to use TensorFlow 2 to build and train convolutional neural networks (CNNs). You’ll begin with a detailed exploration of convolution—what it is, why it matters, and how to integrate it into neural networks. From there, you’ll apply CNNs to a range of image recognition datasets, progressing from simple to complex challenges. You will also learn how to perform text preprocessing and classification with CNNs. Finally, the course covers advanced techniques such as batch normalization, data augmentation, and transfer learning to boost performance in computer vision tasks. By the end, you will have the skills to confidently build and optimize CNNs with TensorFlow for real-world deep learning applications.
Who should take this Course?
The Deep Neural Networks using Python Online Course is ideal for data scientists, machine learning practitioners, software developers, and AI enthusiasts who want to gain hands-on experience in building and training deep learning models with Python. It is also suitable for students, researchers, and professionals in fields like computer vision, natural language processing, and automation who are eager to apply deep neural networks to solve complex real-world problems.
What you will learn
- Learn the basics of machine learning and neural networks
- Understand the architecture of neural networks
- Learn the basics of training a DNN using the Gradient Descent algorithm
- Learn how to implement a complete DNN using NumPy
- Learn to create a complete structure for DNN from scratch using Python
- Work on a project using deep learning for the IRIS dataset
Course outline
Introduction
- Course Promo
- Introduction to Instructor
- Introduction to Course
Basics of Deep Learning
- Problem to Solve Part 1
- Problem to Solve Part 2
- Problem to Solve Part 3
- Linear Equation
- Linear Equation Vectorized
- 3D Feature Space
- N-Dimensional Space
- Theory of Perceptron
- Implementing Basic Perceptron
- Logical Gates for Perceptrons
- Perceptron Training Part 1
- Perceptron Training Part 2
- Learning Rate
- Perceptron Training Part 3
- Perceptron Algorithm
- Coding Perceptron Algo (Data Reading and Visualization)
- Coding Perceptron Algo (Perceptron Step)
- Coding Perceptron Algo (Training Perceptron)
- Coding Perceptron Algo (Visualizing the Results)
- Problem with Linear Solutions
- Solution to Problem
- Error Functions
- Discrete Versus Continuous Error Function
- Sigmoid Function
- Multi-Class Problem
- Problem of Negative Scores
- Need of SoftMax
- Coding SoftMax
- One-Hot Encoding
- Maximum Likelihood Part 1
- Maximum Likelihood Part 2
- Cross Entropy
- Cross Entropy Formulation
- Multi-Class Cross Entropy
- Cross Entropy Implementation
- Sigmoid Function Implementation
- Output Function Implementation
Deep Learning
- Introduction to Gradient Descent
- Convex Functions
- Use of Derivatives
- How Gradient Descent Works
- Gradient Step
- Logistic Regression Algorithm
- Data Visualization and Reading
- Updating Weights in Python
- Implementing Logistic Regression
- Visualization and Results
- Gradient Descent Versus Perceptron
- Linear to Non-Linear Boundaries
- Combining Probabilities
- Weighted Sums
- Neural Network Architecture
- Layers and DEEP Networks
- Multi-Class Classification
- Basics of Feed Forward
- Feed Forward for DEEP Net
- Deep Learning Algo Overview
- Basics of Backpropagation
- Updating Weights
- Chain Rule for Backpropagation
- Sigma Prime
- Data Analysis NN (Neural Networks) Implementation
- One-Hot Encoding (NN Implementation)
- Scaling the Data (NN Implementation)
- Splitting the Data (NN Implementation)
- Helper Functions (NN Implementation)
- Training (NN Implementation)
- Testing (NN Implementation)
Optimizations
- Underfitting vs Overfitting
- Early Stopping
- Quiz
- Solution and Regularization
- L1 and L2 Regularization
- Dropout
- Local Minima Problem
- Random Restart Solution
- Vanishing Gradient Problem
- Other Activation Functions
Final Project
- Final Project Part 1
- Final Project Part 2
- Final Project Part 3
- Final Project Part 4
- Final Project Part 5