Instance-Based vs Model Based
Before Learning what is Instance-based learning and how it differs from model-based learning, we need to understand what is a use case and what role a Machine Learning model plays in this.
Use case
In the context of machine learning (ML), a use case refers to a specific application or scenario where ML techniques are utilized to solve a problem or achieve a specific goal.
A machine learning model is created to fulfill the use cases by training the model or enhancing the pre-built models to fulfill the purpose for the various kinds of use cases.
So, a Machine-Learning Model learns about data in 2 ways, which are
Instance-based Techniques
Model-Based Techniques
To comprehend which techniques are used in a particular ML model or which technique should be used to create a model, we focus on the learning pattern.
Instance-Based ML Technique
Instance-based machine learning is like learning by memorizing. It's like when you try to remember answers to specific questions for a test without really understanding the concepts. In instance-based machine learning, the computer stores the examples it has seen and uses them directly to make predictions or decisions.
For better illustrations, Instance-based machine learning is like a child learning to recognize different types of fruits based on their appearance. Imagine a child being shown pictures of different fruits, such as apples, bananas, and oranges. The child looks at the pictures and remembers the specific fruits based on their shape, color, and texture. When presented with a new fruit, the child compares its features with the stored examples and makes a prediction about what fruit it might be.
Model-Based ML Technique
Model-based machine learning is like understanding the underlying concepts and using them to make predictions. It's like when you really understand the principles of math and can solve different math problems, even if you haven't seen the exact same problem before. In model-based machine learning, the computer tries to understand the patterns or rules in the examples and uses that understanding to make predictions or decisions.
For example, the child might notice that when the sky is clear and the sun is shining, it's usually a warm day. But when dark clouds appear and it starts raining, it's usually a cooler day. The child learns to understand these patterns and creates a mental model or rule that helps them predict the weather based on observable cues.
In the first example, the child is using an instance-based approach, where they memorize specific examples (e.g., pictures of fruits) to make predictions. In the second example, the child is using a model-based approach, where they understand the underlying patterns (e.g., weather cues) and use that understanding to make predictions for new situations.
Conclusion
In summary, instance-based machine learning is like memorizing specific examples for predictions, while model-based machine learning is like understanding patterns or rules and using that understanding to make predictions in new scenarios.
That's the end of the article readers!
Will be explaining more in my following blogs!
Creating a machine learning model is like exploring a new world. You start with random values and predictions, but with each step, you learn from the results and adjust your path. Persistence and accuracy eventually lead to success." - Andrew Ng
Do subscribe and keep supporting! 😊