Classification
- (Duke University - Cheng-Yu Chen)
- Overview
In statistics, classification is the process of determining which category an observation belongs to. For example, classifying an email as spam or not spam, or assigning a diagnosis to a patient based on their characteristics.
Statistical classification is a supervised learning method that trains a program to categorize new information based on its relevance to known data. It's a machine learning method that classifies data into groups based on similarities. This is done by training a classifier on a dataset, which is then used to predict the class of new data.
The most common methods for classification are based on Euclidean spaces. One widely used method is support vector machines(SVM). SVM is based on optimizing the gap in feature space between the training cases in the two classes.
Please refer to the following for more information:
- Wikipedia: Regression Analysis.
- Wikipedia: Correlation Analysis.
- Wikipedia: Data Classification.
- Wikipedia: Decision Tree Learning.
- Decision Trees in Machine Learning
Decision trees in machine learning (ML) provide an effective method for making decisions because they lay out the problem and all the possible outcomes. It enables developers to analyze the possible consequences of a decision, and as an algorithm accesses more data, it can predict outcomes for future data.
In ML, a decision tree is an algorithm that can create both classification and regression models. The decision tree is so named because it starts at the root, like an upside-down tree, and branches off to demonstrate various outcomes.
- Classification Trees
A classification tree is a type of decision tree that is used in machine learning (ML) to classify data as part of a known object class. Classification trees are a type of supervised ML algorithm that uses conditional statements to divide training data into subsets. Each split adds complexity to the model, which can be used to make predictions.
Classification trees are used to predict categorical data, such as whether an email is spam or not. They determine whether an event happened or didn't happen, usually involving a “yes” or “no” outcome.
Classification trees are complex and can lead to overfitting. One way to avoid this is to set a minimum number of training inputs to use on each leaf. For example, you can use a minimum of 10 passengers to reach a decision, and ignore any leaf that takes less than 10 passengers.
- Classification in Machine Learning
Classification is a supervised machine learning method where the model tries to predict the correct label of a given input data. In classification, the model is fully trained using the training data, and then it is evaluated on test data before being used to perform prediction on new unseen data.
Supervised Machine Learning is where you have input variables (x) and an output variable (Y) and you use an algorithm to learn the mapping function from the input to the output Y = f(X). The goal is to approximate the mapping function so well that when you have new input data (x) you can predict the output variables (Y) for that data.
Supervised learning problems can be further grouped into Regression and Classification problems.
- Pattern Recognition and Classification
Here are some steps in PR:
- Identify common elements
- Identify and interpret common differences
- Identify individual elements
- Describe patterns
- Make predictions based on patterns
Feature extraction is a crucial step in pattern recognition. It involves selecting and representing the most relevant information or attributes from the raw data.
Hybrid pattern recognition combines two or more methods to identify patterns. For example, a system may use both statistical and structural methods to identify patterns in medical images.
[More to come ...]
