Liver Disease Classification by Pruning Data Dependency Utilizing Ensemble Learning Based Feature Selection
This repository is the implementation of Liver Disease Classification by Pruning Data Dependency Utilizing Ensemble Learning Based Feature Selection for the AI 2022: Advances in Artificial Intelligence that was presented in 35th Australasian Joint Conference, AI 2022, Perth, WA, Australia. The epdf link is Here.
Liver disease is responsible for over 2 million additional deaths globally each year. Therefore, early detection and treatment may lower the likelihood of liver disease-related death. This study suggests a methodology that integrates approaches for classifying liver disease by reduction of data dependency, which gives the advantage of getting more accurate predictions even with less data. Moreover, Two imputation strategies were employed to tackle missing value and were contrasted with each other. Despite showing slight differences, no statistically significant distinctions between them were found. Machine Learning (ML) methods such as Random Forest, Extra Trees, Support Vector Machine, and K-Nearest Neighbor and neural network such as Multilayer Perceptron were employed to categorize liver diseases.
- Exploratory Data Analysis.
- Attractive visuals.
- Statistical tests for normality, skewness, comparing dataset with two sample t-test
- Deductive Reasoning
- Nested Cross Validation
- Implementation of published paper
- Machine Learning Classification in proper manner and many more
If you use this project by any means, we would be grateful if you cite our paper!
Khaled, M.A.B., Rahman, M.M., Quaiyum, M.G., Akter, S. (2022). Liver Disease Classification by Pruning Data Dependency Utilizing Ensemble Learning Based Feature Selection. In: Aziz, H., Corrêa, D., French, T. (eds) AI 2022: Advances in Artificial Intelligence. AI 2022. Lecture Notes in Computer Science(), vol 13728. Springer, Cham. https://doi.org/10.1007/978-3-031-22695-3_43
For citation file download, Click Here
The liver disease dataset can be found at the link.
This Project is open-sourced software licensed under the MIT license.
