Traditional Programming vs Machine Learning
Summary: Traditional programming is fundamentally different from conventional programming. In Traditional programming, the program uses data in a logical way to achieve a task. Machine learning uses data to arrive at a logic that can further be used to predict patterns .
Here are some examples of conventional programs
- Java or PHP code rendering web pages
- C++ or Java programs running behind ATMs
In all of these class of programs, the programmer codes the logic into the program. Based on the situation (data), the logic weaves through the program to perform some actions.
Machine Learning Program
The program that we have seen in the Machine Learning demo is a good example of a machine learning program. We haven’t coded a fixed logic into the program – instead the program is designed in such a way that the algorithm derives the logic based on the data. Once we got the logic, we can use further data to arrive at the results.
The first set of data is called the training data and we validate that using the test data. Once we fine-tune the logic, we can apply it further on real scenarios.