-
Notifications
You must be signed in to change notification settings - Fork 16
What is Machine Learning ?
Awantik Das edited this page Apr 1, 2017
·
1 revision
Making software learn.
- Categorical Data/ Qualitative information - In forms of strings
- Continues Data - In form of numbers
- Model - The function that take all parameters & predicts the label
- You have data - continues & categorical. And, labels
- Model here uses the data & label information to learn & then predict labels for unlabelled data.
- The data that is available with labels is called training data
- The data that is not avilable with labels is called testing data.
- Two types of supervised machine learning - regression & classification
The labels of the data which are continues in nature falls in this category. Like House price prediction
The labels of the data is categorical in nature falls in this category. Cancer detection.
- Label information is not available here
- There is no training step involved here unlike supervised ml.
- Two types of - clustering & dimensionality reduction.
- Identifies groups or patter in data. Divide students into groups
- Identify the minimal features that can reprsent the dependencies of house price. Tenent info may not matter in house price prediction