Table 2  A summary of the machine learning methods employed in LTx

Support vector machine (SVM)

Support Vector Machine or SVM is one of the most popular classification algorithms for creating the best decision line or boundary. Its objective is to find a hyperplane in N-dimensional space (N is the number of features) that distinctly classifies the data points.

Logistic Regression

Logistic regression is utilized to evaluate the association of independent (predictor) features with a binary dependent (outcome) feature.

Decision Tree

A decision tree uses a set of rules to classify and visualize numerical and categorical data. A Decision Tree is used to generate simple and logical rules.

Random Forests (RF)

A random forest classifier is a meta-estimator that fits many decision tree models under different samples of the data sets. RF employs the average of decision trees to improve the model’s prediction accuracy and control overfitting.

Bayesian network and Naïve Bayes

The Naive Bayes algorithm was developed based on the Bayes theorem assuming independence between each pair of features. This algorithm demands a small amount of training data to estimate the necessary parameters.

Neural Networks

Neural networks or artificial neural networks (ANN) are a type of artificial intelligence that can be used in medicine for early and more accurate diagnosis of diseases. They make it possible to distinguish patients from those who are healthy.

Markov Model

Markov models are often used to model the probabilities of different states and the transition rates between them. This method is generally used to detect patterns, and predict and learn statistics of sequential data.

K-means

The k-Means algorithm is a clustering algorithm used to predict the probability of disease based on medical data sets.

Gradient Boosting trees (XGBoost)

Gradient boosting is a machine learning algorithm where tree-based classifiers are trained to reinforce each other to achieve outstanding outcomes. This method differs from Random Forests (RF) where trees are learned sequentially based on the performance of all previous trees.

Convolutional Neural Network (CNN)

The CNN-based deep neural system is widely used in the medical classification task. CNN is an excellent feature extractor to classify medical images to overcome complicated and expensive feature engineering.

KNN

K-Nearest-Neighbors (KNN) is one of the successful data mining techniques used in classification problems that refers to the number of nearest neighbors.