Pluralsight Course on “Advanced Machine Learning With ENCOG”

Hello All Machine Learning Enthusiasts,

Recently, my new course on machine learning “Advanced Machine Learning With ENCOG ” got published on Pluralsight.com. Although the course is targeted for ENCOG users , but the concepts discussed in this course are widely applicable in other frameworks also.  This advanced level course is segregated in multiple parts.

This first part is available here.

Here is the table of contents for the first part of the course :

Module 1 (Course Introduction ): This module will provide an overview of this course and why one should take this course. This module will also describe the scope of this course. There is also a quick recap clip inside this module, which will help you to brush up the basics before delving into the advanced topics of machine learning.

Module clips are as follows :

  1. Introduction
  2. Course Scope
  3. Course Structure
  4. Quick Recap
  5. Summary

Module 2 (Network Tuning Part 1) : This module will be focused on various network tuning techniques to improve the accuracy of neural network predictive models. Network Tuning module has been divided into two parts. In the Part 1, We will look at network size tuning using pruning process. We will also discuss the need of such tuning. We will take a C# WPF application demo to understand network size optimization and how it can be used to solve the major problems of underfitting and overfitting.

Module clips are as follows :

  1. Introduction
  2. Outline
  3. Network Tuning
  4. UnderFitting And Overfitting
  5. Selection of layers and neurons
  6. Why Network Pruning ?
  7. About Pruning
  8. ENCOG support for Pruning
  9. Training ,Cross Validation and  Test dataset
  10. Demo Introduction
  11. Demo : XAML Code
  12. Demo : Core Steps-Shuffle, Segregate, Normalize And Prune
  13. Demo : Core Steps-Train
  14. Demo : Observations
  15. Summary

Module 3 (Network Tuning part 2) : In the first part, we looked at the requirement of network tuning and we talked in detail about two issues of learning process, first was underfitting and  second was overfitting. And we had discussed the network size optimization using the pruning process to address these two key issues.In this module, we will take another aspect of network tuning, which is to tune the training process itself. We will look at various strategies and how ENCOG machine learning framework can help us to implement these strategies.

Module clips are as follows :

  1. Introduction
  2. Outline
  3. Training Process Tuning
  4. ENCOG Training Strategies
  5. Greedy Strategy
  6. Demo : Greedy Strategy
  7. Hybrid Strategy
  8. Demo : Hybrid Strategy
  9. Reset Strategy
  10. Demo : Reset Strategy
  11. Required Improvement Strategy
  12. Demo : Required Improvement Strategy
  13. Smart Learning Rate and Smart Momentum Strategy
  14. Demo : Smart Learning Rate and Smart Momentum Strategy
  15. StopTrainingStrategy
  16. Demo : Basic Stop Strategies
  17. Demo : StopTrainingStrategy
  18. EarlyStoppingStrategy
  19. Demo : EarlyStoppingStrategy
  20. Summary

Module 4 (Neural Network Architectures Overview) : This module will provide an overall picture of various ENCOG Supported neural network architectures .This module will outline the reasons of co-existence of so many neural network architectures. Finally we will try to map different neural network architectures in different categories using an architectural map.

Module clips are as follows :

  1. Introduction
  2. Outline
  3. Type of Network Covered
  4. Why So Many ?
  5. Architectural Tree
  6. Summary

Module 5 ( Feed Forward Network – Part 1): This is the first part of the Feed forward networks, in which we will learn about linear neural networks. We will start with learning the concept of linear and non-linear problems. Then we will take two linear neural networks, Adaline and Linear Perceptron. We will also learn to implement these linear networks using ENCOG framework.

Module clips are as follows :

  1. Introduction
  2. Outline
  3. Feed forward Networks
  4. Input Output Mapping
  5. Linear Versus Non-Linear” Linear Neural Networks
  6. Adaline Network
  7. Adaline Network in ENCOG
  8. Demo : Adaline Network
  9. Perceptron Network
  10. Perceptron Network in ENCOG
  11. Demo : Perceptron Network
  12. Summary

Module 6 (Feed Forward Network – Part 2) : This is the second part of the Feed forward networks, in which we will learn about non-linear neural networks.We will take two non-linear neural networks, Multi Layer Perceptron and Radial Basis Function Network. We will also learn to implement these non-linear networks using ENCOG framework. We will also look at few of the real world applications of feedforward networks.

Module clips are as follows :

  1. Introduction
  2. Outline
  3. Non-Linear Neural Networks
  4. Multi Layer Perceptron
  5. Multi Layer Perceptron in ENCOG
  6. Demo : MLP Network
  7. RBF (Radial Basis Function) Network
  8. Radial Basis Function Calculation
  9. RBF Network Implementation
  10. XOR Problem using RBF
  11.   Basic RBF Network Implementation
  12. RBF Network in ENCOG
  13. Demo : RBF Network
  14. Applications of Feedforward Networks
  15. Summary

Module 7 (Feedback Networks) : This module will provide an overview of feedback networks and its properties. We will learn in detail about two simple recurrent networks in this module, first  one is the ELMAN network and second is JORDAN network. We will also look at real world applications of feedback networks. We will also take few C# demos to learn the implementation of feedback networks using ENCOG machine learning framework.

Module clips are as follows :

  1. Introduction
  2. Outline
  3. Feedback Network
  4. Elman Network
  5. Temporal XOR Problem
  6. Elman Network Training
  7. Elman Network in ENCOG
  8. Demo : Elman Network
  9. Jordan Network
  10. Jordan Network in ENCOG
  11. Demo : Jordan Network
  12. Applications of Recurrent Networks
  13. Summary

Module 8 (Course Summary) :This module will provide a quick summary to this advanced level course. We will also look at a short glimpse of the next course.

Module clips are as follows :

  1. Introduction
  2. Summary
  3. Next Course Glimpse
As this first part is focused towards supervised networks, the second part will be focused on unsupervised networks.
To all viewers, it would be great if you can give the feedback also, so that the feedback could be incorporated to improve future courses.

Finally, I would also like to mention two great books written by Jeff Heaton. This are great ebooks and must have for all ENCOG users.

  1. Programming Neural Networks with Encog3 in C#, 2nd Edition by Heaton, Jeff (Oct 2, 2011)
  2. Introduction to Neural Networks for C#, 2nd Edition by Jeff Heaton (Oct 2, 2008)
About these ads

Abhishek is a software architect, developer and Pluralsight author. He is very passionate about working with data especially in the field of machine learning. He has authored several courses on machine learning which are available on Pluralsight. He has been involved in several software development projects, which involves various machine learning techniques. His work focuses on architecting and developing applications especially in the area of monitoring, optimization, pattern recognition, and fault detection. His professional interests include software design patterns, agile practices, and various technologies such as WCF, WF, WPF, Silverlight, SQL Server, Entity Framework and ASP.NET MVC. He is also a Microsoft Certified Professional (HTML5, Javascript, CSS3).

Posted in Machine Learning
4 comments on “Pluralsight Course on “Advanced Machine Learning With ENCOG”
  1. Ali Umair says:

    Hi, when your next level course of Advance Machine learning is expected to release ?

  2. Spiky says:

    Hi! I just finished your introduction course on Pluralsight for Encog, and I started the advanced course. I was wondering if you would release another Encog course on Optimization techniques of Encog, or maybe a course that just makes a brief survey of the most important techniques, to help to know which one fits best for a given problem. Thanks!

    • Hi !! First of all, thanks for viewing my course. I hope it helped you a little bit. And Yes, I am coming up with my next course on Pluralsight, in which I will talk about unsupervised learning with self organizing map for dimensionality reduction and clustering, pattern recall and reconstruction using hopfield network, time series forecasting, and optimization using genetic algorithm.
      I have completed this course already, and its under production process.So, I will be live pretty soon. So stay tuned and keep giving feedback.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

About Me


Software Architect, Developer & Pluralsight Author

Software Architect, Developer & Pluralsight Author, Twitter Handle :

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 71 other followers

Disclaimer
The opinions expressed herein are my own personal opinions and do not represent my current or previous employer's view in anyway.
Live Feed
Follow

Get every new post delivered to your Inbox.

Join 71 other followers

%d bloggers like this: