Glossary¶

Accuracy¶

The fraction of predictions a classification model has correctly identified. Formula:

\[ \text{accuracy} = \frac{\text{True Positives + True Negatives}}{\text{Total amount of data points}} \]

See here for respective definitions: True Positives True Negatives

Application Programming Interface (API)¶

An API allows independent applications to communicate with each other and exchange data.

Classification¶

Classification refers to dividing a data set into multiple chunks, which are then considered "classes".

Confusion Matrix¶

A confusion matrix is a table that is used to define the performance of a classification algorithm. It classifies the predictions to be either be true positive, true negative, false positive or false negative.

Decision Tree¶

A Decision Tree represents the process of conditional evaluation in a tree diagram.

Implemented in Safe-DS as DecisionTreeClassifier and DecisionTreeRegressor.

F1-Score¶

The harmonic mean of precision and recall. Formula:

\[ f_1 = 2 \cdot \frac{\text{precision} \cdot \text{recall}}{\text{precision} + \text{recall}} \]

False Negative (FN)¶

An outcome is considered to be a false negative, if the data model has mistakenly predicted a value of negative class.

False Positive (FP)¶

An outcome is considered to be a false positive, if the data model has mistakenly predicted a value of positive class.

Feature¶

Each feature represents a measurable piece of data that can be used for analysis. It is analogous to a column within a table.

Linear Regression¶

Linear Regression is the supervised Machine Learning model in which the model finds the best fit linear line between the independent and dependent variable i.e. it finds the linear relationship between the dependent and independent variable.

Implemented in Safe-DS as [LinearRegressor[safeds.ml.classical.regression.LinearRegressor].

Machine Learning (ML)¶

Machine Learning is a generic term for artificially generating knowledge through experience. To achieve this, one can choose between a variety of model options.

Metric¶

A data metric is an aggregated calculation within a raw dataset.

One Hot Encoder¶

If a column's entries consist of a non-numerical data type, using a One Hot Encoder will create a new column for each different entry, filling it with a '1' in the respective places, '0' otherwise.

Implemented in Safe-DS as OneHotEncoder.

Overfitting¶

Overfitting is a scenario in which a data model is unable to capture the relationship between the input and output variables accurately, due to not generalizing enough.

Positive Class¶

The "Positive Class" consists of all attributes to be considered positive. Consequently, every attribute to not be in this class is considered to be negative class.

Precision¶

The ability of a classification model to identify only the relevant data points. Formula:

\[ \text{precision} = \frac{\text{True Positives}}{\text{True Positives + False Positives}} \]

See here for respective references: True Positives False Positives

Random Forest¶

Random Forest is an ML model that works by generating decision trees at random.

Implemented in Safe-DS as RandomForestClassifier and RandomForestRegressor.

Recall¶

The ability of a classification model to identify all the relevant data points. Formula:

\[ \text{recall} = \frac{\text{True Positives}}{\text{True Positives + False Negatives}} \]

See here for respective references: True Positives False Negatives

Regression¶

Regression refers to the estimation of continuous dependent variables.

Regularization¶

Regularization refers to techniques that are used to calibrate machine learning models in order to minimize the adjusted loss function and prevent overfitting or underfitting.

Sample¶

A sample is a subset of the whole data set. It is analyzed to uncover the meaningful information in the larger data set.

Supervised Learning¶

Supervised Learning is a subcategory of ML. This approach uses algorithms to learn given data. Those Algorithms might be able to find hidden meaning in data - without being told where to look.

Tabular Dataset¶

In addition to a regular table, a tabular dataset will mark one column as "target", meaning that an applied algorithm will train to predict its entries.

Target¶

The target variable of a dataset is the feature of a dataset about which you want to gain a deeper understanding.

Test Set¶

A set of examples used only to assess the performance of a fully-specified classifier.

Training Set¶

A set of examples used for learning, that is to fit the parameters of the classifier.

True Negative (TN)¶

An outcome is considered to be a true negative, if the data model has correctly predicted a value of negative class.

True Positive (TP)¶

An outcome is considered to be a true positive, if the data model has correctly predicted a value of positive class.

Underfitting¶

Underfitting is a scenario in which a data model is unable to capture the relationship between the input and output variables accurately, due to generalizing too much.

Validation Set¶

A set of examples used to the parameters of a classifier.