Machine Learning in Action (I)-Introduction: A First Look at Machine Learning
Unmanned aircraft applications
Supervised learning means that first some dataset needs to be provided and for each data in the dataset, there is a corresponding correct answer and the (training set) algorithm is making predictions based on these correct answers. It is further divided into regression problems and classification problems.
1.Regression problems: predicting a continuous output value by regression.
2.Classification problem: Predicting the output of a discrete value by classification.
Given data about the size of houses on the real estate market, try to predict their price. Price as a function of size is a continuous output, so this is a regression problem. We could turn this example into a classification problem by instead making our output about whether the house "sells for more or less than the asking price." Here we are classifying the houses based on price into two discrete categories.
Is a straight line fit or a quadratic function fit better
(a) Regression - Given a picture of a person, we have to predict their age on the basis of the given picture
(b) Classification - Given a patient with a tumor, we have to predict whether the tumor is malignant or benign.
The input data for unsupervised learning is not labeled and its purpose is to classify the raw material in order to understand the internal structure of the material. It is further divided into clustering algorithms and non-clustering algorithms.
Clustering algorithms: can be used to organize large clusters of computers; can be used for analysis of social networks; can be used for market segmentation; can be used for analysis of astronomical data.
Classification according to genes
Non-clustering algorithms: the problem of cocktail parties.
b,c