Overview of Machine Learning Classifiers Used in Readmission Prediction

Table of Links

Abstract and 1. Introduction

2. Related Works

3. Methodology and 3.1 Data

3.2 Data preprocessing

3.3. Predictive models

4. Evaluation

4.1. Evaluation metrics

4.2. Results and discussion

5. Conclusion and References

3.3. Predictive models

In this study, multiple data mining algorithms and deep learning were employed to create prediction model.

Logistic regression: Logistic regression, a statistical technique, finds extensive use in binary classification tasks, particularly in health sciences studies where the focus lies on disease states (diseased or healthy) and decision-making scenarios (yes or no) [38].

Random forest: Random forest utilizes a structure composed of numerous decision trees, with predictions from each tree combined to forecast the value of a variable [39].

KNN: KNN is a model that leverages the values of the nearest samples in the training data to determine the category or value of a given sample [40].

SVM: Using SVM, data are transformed into a high-dimensional feature space where separating hyperplanes are constructed to maximize the margin between data points and the hyperplane, effectively delineating them into distinct classes [41]. This process facilitates robust classification by ensuring clear boundaries between different classes in the feature space, enhancing the model's ability to generalize to unseen data.

Gaussian Naive Bayes: Naive Bayes is a probabilistic classifier that employs Bayes' theorem to estimate the probability of a given set of features belonging to a specific label. It calculates the conditional probability of event A occurring given the individual probabilities of A and B, as well as the conditional probability of event B. This approach assumes that features are independent [42]. Gaussian Naive Bayes is a variant of the Naive Bayes classifier that assumes features to follow a Gaussian distribution.

By employing these diverse algorithms, this study aims to explore their effectiveness in predicting patient readmission rates and identifying the most suitable approach for the given dataset and research objectives.