1. How do machines “learn”?
Machine learning (ML) is a branch of artificial intelligence (AI) based on the idea that systems can learn from data, identify patterns and make decisions like a human mind with minimal human intervention.
We teach the computer how to program and reprogram itself based on data we feed it. In short, it is a system that grows on its own.
2. Where do we use machine learning?
Machine learning applications comes big and small, ranging from simple Google search suggestions to self driving vehicles. Online recommendations such as Amazon and Netflix are everyday applications of machine learning. Users feed in personal data, in which the system processes and analyzes to make unique suggestions based on shopping activity. More important uses of ML include fraud detection, where security systems identify unusual behavior in codes, emails, or data.
3. Machine Learning vs AI
While the two sound similar, artificial intelligence is the broader scientific area of mimicking human behavior, while machine learning focuses on training machines how to learn based on data. ML can be considered a subset of AI, more specific on the training and testing methods for development.
4. Good machine learning systems
While machine learning systems sound powerful, they are only as good and accurate as we make them. As they say,
Good ML systems require thorough:
- data preparation,
- good algorithms — both basic and advance,
- iterative process,
- scalability,
- and ensemble modeling, a process creating several models to predict an outcome.
