The Art of Learning

Neural Networks and Education

Francis T.S. Yu, Edward H. Yu, Ann G. Yu

• Publisher: CRC Press
• Year: 2018
• eBook ISBN: 9781351116329
• Edition: 1
• Topic: Computer Science
• Subtopic: Electrical Engineering & Telecommunications
• Index: Computer Science

This book presents the idea that innovative ways of teaching and learning are very essential to retention and growth. Presented in 15 sections, the book starts with the common sense training on education and moves on to neural network operation. Throughout the book, the art of learning, associative, cognitive, and creative learning are stated and defined. Learning simplicity, information content as related to neural network learning are discussed. The author also discusses neural plasticity and adaptability in smarter neural networks.

If we know our human brain's basic abilities and limitation then a better educational methods can be implemented.

• 
  
  

• Presents the idea that innovative ways of teaching and learning are very essential to retention and growth

• 
  
  

• Discusses major differences and constraints between neural network and computer

• 
  
  

• Presents the significances of learning simplicity and information content as related to neural network learning are included

• 
  
  

• Stresses the neural network learning capabilities and limitations and their role in developing more efficient learning techniques

CHAPTER 1

Common Sense Training

Let me start with some common sense in education. In Figure 1, we see an Asian water buffalo and in Figure 2 an Arabian horse. When we try to project for the apparent strengths and limitations between the two in pictures, we see that the water buffalo looks to be very well suited for work in a muddy rice field and might also be a very good swimmer, whereas, the Arabian horse looks very good for riding and could probably be trained to become an excellent racehorse. Common sense, then, would indicate that it would be best to train an Arabian horse for racing, and an Asian water buffalo to work in a muddy field. To do otherwise would be a great mistake!

Inherent Limitations and Abilities of Species

FIGURE  1  Asian Water Buffalo

–Excellent for working in a muddy rice field

–Good swimmer

FIGURE  2  Arabian Horse

–Excellent racing

–Good for riding

Is one better than the other? Common sense tells us we must understand the inherent advantages of each one.

CHAPTER 2

Audio vs. Visual Learning

I have always struggled with language. While I was born in a foreign land and English is my second language, I seem to have a more difficult time with it than my wife, who is also not a native English speaker. She excels at picking up the nuances of the spoken word, is a talented writer, can be detail oriented and has the amazing ability to repeat conversations verbatim. I, on the other hand, often appear not to be listening closely. I like to think visually, instead of verbally and enjoy figuring out how things work. Where language is concerned we are different: she is an Arabian horse and I am a water buffalo (or vice versa!).

Individuals vary in the way they process information. One key difference in learning styles is sequential vs. spatial learning.

Sequential learners are able to listen skillfully and navigate verbal instructions quite easily, learning in a step-like manner. Often they are great storytellers and can mimic sounds and conversations easily. Due to this auditory skill, they are known as Auditory-Sequential Learners. These people excel at all things verbally related and are often talented writers.

Spatial learners are visually attuned, often excelling at pattern recognition and puzzling out the overall picture. Due to their visual acuity, they are known as Visual-Spatial Learners.

There are advantages to both learning styles and most people possess a little of each quality while skewing more towards one or the other. Unfortunately, in an effort to streamline the learning process, education has historically seemed to favor the Auditory-Sequential Learners. Why? A.S. learners are good listeners and are able to navigate oral instructions quite aptly. As a result, they are “easier” to teach. School tends to be a positive experience for them. They are able to express themselves well, are good listeners, learn sequentially, are rapid processors, and think verbally.¹

Visual-Spatial Learners learn via imagery of the whole concept, are spatially aware, astute observers, think holistically, and need more time to process information.²

It can sometimes take V.S. learners longer to express what they are learning and, as a result, they may sometimes seem to be lacking. For them, early academic life is often a negative experience. At least, as a youngster, it was that way for me. In fact, it wasn’t until my last year in high school that I started to recognize my strengths.

Fortunately, the situation is improving. Today there is a better recognition of different learning styles (Figure 3), including of those on the autism spectrum.

FIGURE  3  Distribution of students with strongly preferred learning styles in the regular classroom.

Note. From the visual-spatial resource website (http:/www.visualspatial.org). Adapted with permission.

I would also like to introduce a third style of learning called the Narrative Learner. The Narrative Learner is someone who requires a story to make sense of our world. That story could be told in images or in words and—because of its sophisticated nature—often requires both. Since a story is usually expressed in words, this is where Auditory-Sequential Learners can excel as they have an ear for the sounds and cadences of language. Also of importance in this style is its sequential nature. Time is an important factor in its expression.

All this is to say that we must recognize the inherent strengths and weaknesses of all types of learning styles whether it be auditory-sequential, visual-spatial, or narrative. Remember, the water buffalo is not an Arabian horse!

Notes

1    Linda Kreger Silverman, Ph.D. Identifying Visual-Spatial and Auditory-Sequential Learners: A Validation Study.

2    Ibid.

CHAPTER 3

Human Brains vs. Machines

As I hinted earlier, a fundamental problem with our educational system lies in the narrow focus it makes on transferring a massive amount of information from an authority figure (namely the teacher) to a subject (namely the student), while granting little to no encouragement, or even time, for critical examination of the information. In practice, the problem often lies in making students memorize large amounts of information in a relatively short period of time, without regard to any understanding of where the information might have come from (i.e. who procured it and whose interests might be served by its dissemination), why exactly they should memorize it, what remains unknown and unexplored, or simply gets overlooked or ignored in the dissemination process, and how they themselves might make their own discoveries and come to their own conclusions. In short, a conventional education often treats human beings more like mindless machines than sensing and feeling organisms capable of thinking, wondering, reasoning, creating, and imagining.

Obviously, we are different from machines, and to ignore this fact is not only to render education inefficient in the cultivation of critical thought, reasoning, creativity, and imagination, it is, in a sense, a misuse or even abuse of our humanity.

If we force humans to perform machine-like tasks, we not only risk being far less efficient, accurate, and precise in accomplishing those tasks, but, more importantly, risk dehumanizing the human being. Conversely, if we expect even the most sophisticated machines to behave like humans we are entering the world of science fiction or fantasy. Talk to any competent scientist or engineer who works in robotics, pattern recognition, and artificial neural networks, for example, and she will tell you that even the most high-tech devices currently available cannot come remotely close to matching people in performing our most basic functions. These differences are why, even today, most robots can only perform the most rudimentary and unvarying movements compared to humans. They are able to walk and talk in a primitive fashion, but only within a minimal amount of variation and degree of freedom.

Now let us look at the major differences between neural networks and digital computers. A digital computer has at least one, and sometimes several, central processing units, while a neural network has many (i.e. neurons). Digital computers use sequential processing, while neural networks use parallel processing. Digital computers use programs to execute, while neural networks use associative learning rules.

Therefore, digital computers are very good at computational processing while neural networks are good at cognitive operations such as pattern recognition, understanding a language, etc.

In terms of calculations such as addition, subtraction, multiplication, and division, a computer is more efficient than a human brain

But for cognitive types of operations such as understanding a language, pattern recognition, or thinking, a three-year-old child, or even a dog, can perform b...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Preface
• Chapter 1 ▪ Common Sense Training
• Chapter 2 ▪ Audio vs. Visual Learning
• Chapter 3 ▪ Human Brains vs. Machines
• Chapter 4 ▪ Simplicity in Learning
• Chapter 5 ▪ Memory Tricks
• Chapter 6 ▪ Talent
• Chapter 7 ▪ Curiosity and Discovery
• Chapter 8 ▪ Unsupervised Learning
• Chapter 9 ▪ Mistakes
• Chapter 10 ▪ Ask Questions
• Chapter 11 ▪ Approximation
• Chapter 12 ▪ Every Problem Has Multiple Solutions
• Chapter 13 ▪ The Illusion of Mastery Fast vs. Slow Learning
• Chapter 14 ▪ Late Bloomers
• Chapter 15 ▪ Leadership in Education
• Chapter 16 ▪ A Short Biography A Little Bit about Myself
• Appendix
• Bibliography
• About the Authors