Skip to main content

Ready for the Modern Age with Machine Learning

· 3 min read
Kobkrit Viriyayudhakorn
Kobkrit Viriyayudhakorn
CEO @ iApp Technology

iApp Machine Learning

When we watch clips on YouTube or various movie Streaming platforms, and content that matches our preferences pops up, many people may already know that this is the capability of AI. But what is behind AI's ability to guess so well is Machine Learning. Today, we will get to know more about Machine Learning.

What is Machine Learning?​

Machine Learning is allowing computers to learn by themselves, based on the data we feed them, to find answers or various results.

To illustrate, Machine Learning is like a brain that helps analyze AI.

In traditional programming, we often set rules and boundaries for what AI can do. Therefore, the resulting AI will not do anything beyond the defined commands. But as time goes by and the system becomes more complex, adding more data will make the program unstable.

Instead of setting fixed rules, feeding data and letting the computer learn by itself is easier. And this is where Machine Learning helps solve the problem.

How Many Types of Machine Learning Are There?​

Basically, Machine Learning can be divided into 3 types:

1. Supervised Learning​

This involves feeding data that we want the computer to use for prediction or analysis to get results.

Example:

  • Suppose we want the computer to distinguish between roses and marigolds.
  • The first step is to let the computer know both flowers by inputting data that describes the characteristics of both flowers, along with telling it which flower is a rose and which is a marigold.

In addition to being used for classification, Supervised Learning is also used with continuous data, such as predicting stock prices, which requires using several pieces of information before buying and selling.

2. Unsupervised Learning​

This is the opposite of Supervised Learning because we don't tell the computer that this is a rose and that is a marigold. Instead, we feed the data and let the computer sort it out itself.

Example:

  • Unsupervised Learning will use separation of data by size, color, or structure of the flower.

Limitations:

  • If the data is small and too similar, Machine Learning cannot learn and distinguish.
  • This type is therefore suitable for use with large and clearly different datasets, such as use in marketing.

3. Reinforcement Learning​

Reinforcement Learning is the type that AI learns most like humans because it relies on experience and trial and error to achieve satisfactory results in the future.

Example:

  • Self-driving vehicles: Machine Learning must learn from the surrounding environment, such as road conditions, weather, and traffic.
  • Learning: It must learn how fast to go, whether the route has traffic congestion, to make the journey safer.

Reinforcement Learning helps AI learn from experience and adapt to make the next journey safer.

Summary​

Machine Learning is a technology that allows AI to learn and develop itself from the input data. Currently, it is widely used in data classification, result prediction, and adaptation in complex systems.

This technology continues to grow and develop rapidly, which will open up new opportunities for using AI in various industries in the future.