Logistic Regression 편집하기

'''Logistic regression''' is a statistical and machine learning algorithm used for binary classification tasks, where the output variable is categorical and typically represents two classes (e.g., yes/no, spam/not spam, fraud/not fraud). Despite its name, Logistic Regression is a classification algorithm, not a regression algorithm, as it predicts probabilities of classes rather than continuous values.

==How It Works==
Logistic Regression models the probability of a binary outcome using a logistic function, also known as the sigmoid function. The sigmoid function compresses values to range between 0 and 1, representing the probability of belonging to a particular class. The model predicts the probability that the input belongs to the positive class (1) and classifies it by applying a threshold, often 0.5.

The logistic function is represented by:

P(y=1 | X) = 1 / (1 + e<sup>-(b0 + b1X1 + b2X2 + ... + bnXn)</sup>)

where:
*'''P(y=1 | X)''' is the probability of the output being 1 given the input features.
*'''X1, X2, ..., Xn''' are the input features.
*'''b0''' is the intercept, and '''b1, b2, ..., bn''' are the coefficients of the features.
==Types of Logistic Regression==
*'''Binary Logistic Regression''': Used for binary classification with two possible outcomes (e.g., yes/no).
*'''Multinomial Logistic Regression''': Used when the outcome variable has more than two categories without any ordering (e.g., classifying types of animals).
*'''Ordinal Logistic Regression''': Used when the outcome variable has ordered categories (e.g., ranking levels from low to high).
==Applications of Logistic Regression==
Logistic Regression is widely used across industries due to its simplicity, interpretability, and effectiveness in binary classification tasks:
*'''Healthcare''': Predicting disease outcomes, risk assessments, and patient survival chances.
*'''Finance''': Credit scoring, fraud detection, and risk analysis.
*'''Marketing''': Customer churn prediction, targeting potential buyers, and lead qualification.
*'''Social Sciences''': Survey analysis, where responses fall into categories like agree/disagree or support/oppose.
==Key Metrics for Evaluating Logistic Regression==
To assess the performance of a Logistic Regression model, common metrics include:
*'''[[Accuracy]]''': The proportion of correct predictions.
*'''[[Precision]]''': The ratio of true positive predictions to all positive predictions.
*'''[[Recall]]''': The ratio of true positive predictions to all actual positives.
*'''[[F1 Score]]''': The harmonic mean of precision and recall, useful when dealing with imbalanced data.
*'''[[AUC]]-[[ROC Curve]]''': Measures the model’s ability to distinguish between classes, where a higher Area Under the Curve (AUC) indicates better performance.
==Assumptions of Logistic Regression==
Logistic Regression relies on several assumptions for accurate results:

1. '''Linearity of Independent Variables and Log-Odds''': Assumes a linear relationship between the log-odds of the outcome and the independent variables.

2. '''Independence of Observations''': Observations should be independent of each other to avoid biased results.

3. '''No Multicollinearity''': Independent variables should not be highly correlated with each other, which can be checked using Variance Inflation Factor (VIF).

4. '''Sufficient Sample Size''': Logistic Regression requires a large enough sample size, especially for categorical variables, to make accurate predictions.
==Handling Limitations==
Logistic Regression may not perform well if the relationship between variables is highly non-linear. In such cases, transformations, polynomial features, or using a more complex model like Decision Trees or Neural Networks can be considered.
==See Also==
*[[Linear Regression]]
*[[Support Vector Machine]]
*[[K-Nearest Neighbor]]
*[[Decision Tree]]
*[[Naive Bayes]]