Authors :
Mba Obasi Odim; Uchechukwu Frederick Ekpendu; Bosede Oyenike Oguntunde; Adeniyi Samson Onanaye
Volume/Issue :
Volume 8 - 2023, Issue 5 - May
Google Scholar :
https://bit.ly/3TmGbDi
Scribd :
https://tinyurl.com/2p9t3pc9
DOI :
https://doi.org/10.5281/zenodo.8037024
Abstract :
Investigating the mortality/survival chances
of Viral Hepatitis and Hepatocellular Carcinoma (HCC)
patients could provide informed knowledge for planning
and implementation of efficient and effective strategies
for curtailing the mortality rate of the disease and at the
same time providing more information about the
relationship between HBV/HCV and HCC. This study,
modelled and assessed the performance of some selected
machine learning algorithms (Artificial Neural Networks
(ANN), Decision Tree, K-nearest neighbours (K-NN)
Logistic Regression, Naïve Bayes, Random Forest and
Support Vector Machine (SVM) for the prediction of the
mortality/survival chances of HCC and Hepatitis
patients. The data were collective from UCI machine
learning repository, consisted of clinical test result of 155
hepatitis patients of 20 attributes with 123 survived
patient and 32 mortalities. There were 13 instances with
missing values, which was removed while cleaning the
dataset leaving 142 instances with 116 survivors’ class
and 26 death class. The HCC dataset contained 165
instances with 50 attributes, 102 survivals and 63 death
instances. The algorithms were deployed within the
WEKA environment and the findings revealed that the
Support Vector Machine recorded the highest
classification performance on the both datasets. This was
followed respectively by the Naïve Bayes on the Hepatitis
and the Random Forest on the Hepatocellular
carcinoma. The Decision Tree recorded the least
accuracies on both datasets. The result therefore
suggests that the Support Vector machine, could be a
most appropriate algorithm for developing a
classification system for survival of Hepatitis and
Hepatocellular carcinoma. Hoverer, the performance of
these algorithms could as well be improved with more
dataset.
Keywords :
Machine Learning, Viral Hepatitis, Hepatocellular Carcinoma, Patients, Survival chances.
Investigating the mortality/survival chances
of Viral Hepatitis and Hepatocellular Carcinoma (HCC)
patients could provide informed knowledge for planning
and implementation of efficient and effective strategies
for curtailing the mortality rate of the disease and at the
same time providing more information about the
relationship between HBV/HCV and HCC. This study,
modelled and assessed the performance of some selected
machine learning algorithms (Artificial Neural Networks
(ANN), Decision Tree, K-nearest neighbours (K-NN)
Logistic Regression, Naïve Bayes, Random Forest and
Support Vector Machine (SVM) for the prediction of the
mortality/survival chances of HCC and Hepatitis
patients. The data were collective from UCI machine
learning repository, consisted of clinical test result of 155
hepatitis patients of 20 attributes with 123 survived
patient and 32 mortalities. There were 13 instances with
missing values, which was removed while cleaning the
dataset leaving 142 instances with 116 survivors’ class
and 26 death class. The HCC dataset contained 165
instances with 50 attributes, 102 survivals and 63 death
instances. The algorithms were deployed within the
WEKA environment and the findings revealed that the
Support Vector Machine recorded the highest
classification performance on the both datasets. This was
followed respectively by the Naïve Bayes on the Hepatitis
and the Random Forest on the Hepatocellular
carcinoma. The Decision Tree recorded the least
accuracies on both datasets. The result therefore
suggests that the Support Vector machine, could be a
most appropriate algorithm for developing a
classification system for survival of Hepatitis and
Hepatocellular carcinoma. Hoverer, the performance of
these algorithms could as well be improved with more
dataset.
Keywords :
Machine Learning, Viral Hepatitis, Hepatocellular Carcinoma, Patients, Survival chances.