Risingbd Online Bangla News Portal

Dhaka     Sunday   25 July 2021

Machine Learning Vs. Human Learning

Dr. Md. Altaf Hossain || risingbd.com

Published: 22:30, 31 December 2020   Update: 23:13, 31 December 2020
Machine Learning Vs. Human Learning

Machine learning, in recent years, is a popular technology, which can be used to perform much complex work in the real world.

From ancient times, human beings are striving to develop machine technology by investing their time and passion. Therefore, in this modern era machine keeps an important place among various technology. Industries and peoples are incorporating many technologies into machines to make it faster, automatic, intelligent, and highly functional. In today’s world, many of the human work is replaced by machines and its application are increasing day by day, especially developed countries are integrating this technology with high competition.

Considering the human time, value, and cost, the machine is more preferable in performing many tasks. In addition, a machine can accomplish the task in due time, faster, and accurately.

Now let’s explain the machine. What is a machine? A machine is a human-made device that can perform a task in full or semi-automatic. For example, a vehicle is a type of machine that can move from one place to another place by its own function. A mobile is a machine that can exchange our information from person to person or one place to another place. Now using the mobile device we not only exchange our information but also perform many of our daily life tasks through various applications such as photo, video, sound editing, reading books, pdf, watching videos, play games, and so on. These technologies did not arise in our life abruptly, rather it has arrived step by step through many research, development, and innovation ideas from many scientists, industries, engineers, and so on.

Now let’s learn about machine learning. Machine learning is a technology where machines are trained through algorithms to perform a task. After learning the task, the machine will keep the method in its memory, and perform the same task perfectly whenever required. We can understand this metaphor through some examples.

Example 1) If we train a machine to learn the voice of a dog. After learning the voice of the dog, the program and algorithm will be set in the machine’s memory. Now if we send some animal’s voice for recognition, this machine can recognize the voice is either from a dog or not. If the voice is from the dog, it can also represent the probability of this voice from the dog. Now if the machine is trained to recognize the animal through recognition of its legs, ear, tail in addition to its voice. This machine can also recognize what kind of animal this is from the image of various animals.

Example 2) Let’s say you have an air conditioner (AC) in your house. Let imagine that today temperature is 32 degree and you are setting the AC temperature according to environment temperature today. Tomorrow is another temperature and you going to set AC temperature according to its environment. Now you think that if there is a function where the AC can act and adjust its temperature according to its environment. Machine learning can help in this case, it can be trained in different ways: a) train it to adjust the temperature according to the environment or b) train it to remember your comfortable settings and act in the future as your comfortable mode. Mostly we perform this task through AC’s auto function mode, and it is adjusting the temperature according to the environmental situation. In machine learning, this term is called reinforcement learning that is reacting to an environment.

Machine learning is classified into three categories. 1) Supervised, 2) Unsupervised, and 3) reinforcement learning. The machine is the most popular, wide, and applied technology in recent years due to the increasing amount of information such as text, images, and videos. Its demand and application are increasing rapidly with its sub-branches. From business analytics to robotics, everywhere you can find the application of machine learning. We can learn more about machine learning in another day or you can find more information through searching machine learning using a search engine. The summary of machine learning from my explanation is that a machine performs exactly after it is trained. In simple words, we can say it acts from its previous learning process. This machine learning technology is the product of human research.

Now we move to the term of human learning. Readers may get some basic ideas about machine and machine learning from the above section and may have some predictions about the intention of my article. Human learning is a continuous process of humans, where we start from born and end to death. We start to learn from our families such as parents, siblings, and other relatives. Then, we learn from the school such as teachers, classmates, and seniors. If we believe in religion, we also learn from mosques, temples, churches, and so on. In everywhere we learn a common thing such as morality, humanity, behavior, character, and honesty. But not all we can perform the same as we learn. Many of us not follow what we have learned to act and try to perform the wrong stuff. Hence many unexpected circumstances are arriving in our society. As human beings sometimes we definitely know it is wrong to do, but still, we insist on doing this wrong task. Many of us are doing the wrong thing by intention or without intention. Even we accept the term that ‘it is normal to do wrong as human’. However human-made machine do not perform wrong in its operation except it has some malfunction in its program or body. Machine properly perform the task that it learned before. Maybe it has no biological life, so it does not do the wrong stuff. Even not willing to do wrong. This is the difference between machine learning and human learning.

2020 year with much history in human life has been finished, and 2021 is started. Many firms also present their income, expenditure, and development during their one year cycle. Personally, we also should need to observe our previous one year experience, development, failures. From the knowledge of the previous year, we should have a good plan from the starting of the New Year. We can follow the machine learning technology for our better performance in the present and future. You can segment your problems in smaller and try to solve them step by step. Maybe it is not possible to solve all the problems. But at least try to solve 1-2 problems. Then together we can do better and have a stable, and peaceful society. Finally happy New Year (2021) to all. Wish we will have a better year than before and free from all-natural disaster like COVID-19.

(Dr. Md. Altab Hossin, Researcher, University of Electronic Science and Technology of China. He is also the founder, smart Chinese Teacher (cnpinyin.com). He can be reached at email: [email protected])

Dhaka/Altab Hossin/Nasim