Deep Learning with Python: Optimizing Deep Learning Models

2h 1mIntermediate2025-02-24

Authors

Frederick Nwanganga

Information Technology Professional and Teacher

Course details

Discover techniques to optimize deep learning models by improving their performance and efficiency. Emphasizing practical applications, instructor Frederick Nwanganga guides you through hands-on coding exercises, covering the essentials of data preprocessing and augmentation, regularization methods to minimize overfitting, optimization algorithms, advanced hyperparameter tuning methods, and more.

Learning objectives
Understand the various techniques used to optimize deep learning models.
Preprocess and augment both text and image data.
Leverage Python to implement regularization techniques in a deep learning model to minimize overfitting.
Understand how different optimization algorithms work and when best to use them.
Use Python to tune the hyperparameters of a deep learning model.

Skills covered

Neural Networks and Deep LearningPythonArtificial Intelligence (AI)Programming LanguagesOpen SourceSoftware DevelopmentOne-Off

Concepts

0. Introduction

01 - Optimizing deep learning models
02 - What you should know
03 - Using the exercise files

1. Optimizing Deep Learning Models

04 - The importance of optimizing deep learning models

2. Regularization Techniques

05 - The bias-variance trade-off
06 - Lasso and ridge regularization
07 - Applying L1 regularization to a deep learning model
08 - Applying L2 regularization to a deep learning model
09 - Elastic Net regularization
10 - Dropout regularization
11 - Applying dropout regularization to a deep learning model

3. Loss Functions and Optimization Algorithms

12 - Common loss functions in deep learning
13 - Batch gradient descent
14 - Stochastic gradient descent (SGD)
15 - Mini-batch gradient descent
16 - Adaptive Gradient Algorithm (AdaGrad)
17 - Root Mean Square Propagation (RMSProp)
18 - Adaptive Delta (AdaDelta)
19 - Adaptive Moment Estimation (Adam)

4. Hyperparameter Tuning Techniques

20 - Parameters versus hyperparameters
21 - Key hyperparameters in deep learning
22 - Methods for hyperparameter tuning
23 - Defining a tunable deep learning model in Keras
24 - Using KerasTuner for hyperparameter tuning

5. Advanced Training Techniques

25 - Batch normalization
26 - Applying batch normalization to a deep learning model
27 - Gradient clipping
28 - Applying gradient clipping to a deep learning model
29 - Early stopping and checkpointing
30 - Learning rate scheduling
31 - Training a deep learning model using callbacks

Conclusion

32 - Continuing to optimize deep learning models