Machine Learning Week 1
Machine Learning Machine Learning develops algorithms for making predictions from data Part of Statistics
Machine Learning
Data Data consists of data instances Data instances are represented as feature vectors 180 70 120 80 110 90 Features are chosen for a specific task at hand (Feature Engineering)
Machine Learning is Generalization of a specific task Making predictions about new data instances - Data A consists of 26 coherent groups This data instance belongs to group #18.
Machine Learning consists of Classification Clustering Regression
Classification Training phase - Input: data instances and their true labels -output: the classification model” or “classifier” Testing Phase - Input: a data instance - output: Its label
Example Systolic BP Negative instances Positive instances HR
K-Nearest-Neighbors Classifier Systolic BP Negative instances Positive instances HR
Support Vector Machine (SVM) Systolic BP HR
SVM can be Nonlinear separable classifier Systolic BP HR
SVM can be multi-class classifier Systolic BP HR
Decision Trees 0 1 2 3 4 5 6 7 8 9