Strategic Analysis in Prediction of Liver Disease Using Different Classification Algorithms

Introduction

Healthcare is regarded as a significant component in enhancing the health-related services for every individual. It makes a provision to improve the health by taking the certain essential measures into consideration and mainly deals with the enhancement of health through the diagnosis of diseases at the right time (Saritha et al., 2017). Thus, “The main aim behind the Healthcare System is to deliver the best quality of services and to predict the diseases at an early stage.”

Liver is an imperative organ of our body. There is a great need for an early detection of liver disease so as to prevent complete liver failure, which can result in patient’s death. For the proper diagnosis, it is necessary to evaluate some of the main attributes of liver patient’s dataset (Vijayarani et al., 2015). Some of the main attributes of liver disease include, “Total_bilirubin, direct_bilirubin, alkaline_phosphotas, total_protein, albumin and globulin_ratio.” Below, Figure 1 shows the various functions that are performed by the liver that makes it the second largest organ in our body.

Figure 1.

Functions performed by the liver

It is a rigorous task for doctors to detect the liver disease accurately. Various classifiers have been utilized to classify the data and to predict the disease through the liver patient’s dataset (Ghosh et al., 2017). “Having access to classification algorithms with huge volume of data will help clinicians to come up with optimal decisions and ultimately improve the overall experience of the patient.” This paper exhibits a survey about the techniques that can be utilized to reveal the disease and gives a roadmap for future work, such as which classification technique to be utilized further for diagnosis of the liver disease.

A. Tool Used

Weka Tool

Weka is an efficient machine learning software that is widely used to classify various parameters when different algorithmic approaches are applied based on the datasets. It is an accumulation of tools utilized for the purpose of visualization and algorithms for analyzing the data and predictive modelling. Classification algorithms are implemented using Weka that are utilized for an early stage detection of disease (Pathan et al., 2018). The experimental results of classification algorithms provide ease to the doctors through the provision of accurate patient’s pathological status. As shown in Figure 2, Weka can be utilized to perform several data mining tasks.

Figure 2.

Data Mining Tasks performed by Weka

Orange Tool

Orange is a data mining tool that constitutes several components. It includes various widgets that are utilized for data analysis and visualization. In this tool, Python is used as a scripting language. It consists of several components for data preprocessing, modeling, ranking, test and scoring. It is compatible with Python, C and C++. Various features are utilized for data analytics. Also, it is an open source machine learning software build on Python. It has a better debugger than the other tools. Moreover, it can be efficiently utilized for comparison of several classification algorithms by analyzing various parameters such as accuracy and precision values. Further, this tool can be utilized to maintain the huge amount of data and to identify various patterns.