What is machine learning?
- Core Task
- Finds patterns in data
- Use those patterns to predict future
- What is learning?
- Identifying patterns
- Recognizing those patterns when you see them again
Machine can find a pattern in existing data, then create and use a model that recognize those patterns in new data
A Data Scientist needs to master
- statistics
- machine learning software
- some problem domain (application fields)
Languages
- R
- Python
Steps of Sovling a Problem
- Ask a question first? such as…
- How to keep a player to play longer?
- How to get a player to play instantly?
Whether or not having lots of right data?
Pattern, Model and Application?
- Need knowledge:
- Training data: the prepared data used to create a model
- Supervised learning: the value you want to predict is in the training data, data is labelled, the target value is part of training data.
(and unsupervised learning) - Classifying machine learning problems and algorithms
- Supervised: Regression, Classification,
- Unsupervised: Clustering Algorithms: Decision tree, Neural network, Bayesian, K-means
- Training a model
- Testing a model
- Using a model