Data mining and data visualization for analysing the rate of bed availability at hospitals due to COVID 19

This study started in July 2020 during the COVID 19 pandemic period to analyze & visually illustrate insights of data of biomedical facilities’ information. The objective of this study is to present major issues faced, solutions found, and a roadmap for future work in developing visual analytics for interactive & data visualization for biomedical facility applications. This chapter starts with a brief introduction of data mining and data visualization, followed by a description of data mining tasks and we’ll build a decision tree model, followed with the concrete examples on various data visualization charts(namely Histogram, Bar Chart, Pie Chart, Line graph, etc), this paper which is intended for visual analysis of “Bed availability rate” in hospitals and decision support for patients and is based on Data Mining & Visualization techniques. Research Article Data mining and data visualization for analysing the rate of bed availability at hospitals due to COVID 19 Thanuj Kumar S1*, Vinitha Dominic K2 and Sumathi V1 1Pursing Computer Science and Engineering, Presidency University, Bengaluru, India 2Vinitha Dominic K, Assistant Professor, Department of Computer Science and Engineering, Presidency University, Bengaluru, India Received: 22 February, 2021 Accepted: 17 March, 2021 Published: 18 March, 2021 *Corresponding author: Thanuj Kumar S, Pursing Computer Science and Engineering, Presidency University, Bengaluru, India, E-mail:


Introduction
In recent times demand has risen for patients' health affected by a coronavirus, so that healthcare should be delivered at the appropriate health center, by the appropriate analysis, with the appropriate means at the level of individual patients to fi ght against the widespread of coronavirus. This study aims to provide data integration across heterogeneous biomedical facilities (like no. of beds available) information to facilitate improved data insights, visual analysis and ultimately to be stepping stone for future trends in this study.
Data stored in large databases are not always comprehendible by human beings, it needs to be cleaned, sorted, and analyzed fi rst, stored records are raw amounts of data-poor in accurate insightful information, not only because is it abundant and seamlessly irrelevant but also continuously increasing, updating and changing. Here is where data mining and visualization comes into the picture. Data mining and visualizations are knowledge discovery tools used for the analysis of data stored in large sets in many different ways as these large data cannot manually predict the outcomes mining tools and visualization techniques provide automated means to comprehend such continuously changing data sets. Data mining is defi ned as the automated process of fi nding patterns, future results, and trends in the data set. On the other side, data visualization is the process of visually communicating and representing the data to bring insights out of it. Human brain able to understand and comprehend visual graphics more easily than numbers and letters. Human brains can absorb graphs, charts, and models quicker than digits in text format. Visualization of such graphical data helps the human brain fi gure out the insights and recognize such knowledge hidden in the data. The goal of data visualization is to not only summarize the abundant dataset but also provide a better way of exploring the knowledge hidden and quickly perceive the insights autonomously [1][2][3][4][5][6][7][8][9][10].
In the next section, we shall demonstrate the Data Visualization Methods to visually analyze the availability of major hospital facility i.e Beds by using various Data Visualization charts

Data mining tasks
The data mining undertakings can be arranged for the Citation: Thanuj Kumar S, Vinitha Dominic K, Sumathi V (2021) Data mining and data visualization for analysing the rate of bed availability at hospitals due to COVID -19. Arch Biomed Sci Eng 7(1): 001-004. DOI: https://dx.doi.org/10.17352/abse.000023 most part into two sorts dependent on what a particular assignment attempts to accomplish. Those two classifi cations are descriptive tasks and predictive tasks. The descriptive data mining task describes the overall properties of information while predictive data mining tasks perform derivation on the accessible informational collection to foresee how another informational index will carry on.
There are different data mining tasks, for instance, game plan, desire, time-course of action assessment, association, clustering, overview, etc. All of these tasks are either predictive data mining assignments or enchanting data mining tasks. A data mining system can execute at any rate one of the abovedemonstrated undertakings as a signifi cant part of data mining.
In this work, we are utilizing Decision Tree Classifi cation for Patients, which helps in choosing to get admitted to a correct emergency clinic dependent on the accessibility of beds.

Bar graph
A bar outline or bar diagram is a graph or chart that presents all-out information with rectangular bars with statures or lengths corresponding to the qualities that they speak to. The bars can be plotted vertically or on a level plane.

Result and conclusion
We describe the results of our analysis, which showed the rate of bed availability in hospitals based on their other related attributes.
We started with the Decision tree which plays a vital role in making decisions, as a result, which helps in predicting the right hospital which has a high rate of bed availability for patients to get admitted. We then demonstrated data visualization with various charts to visually analyze the status of hospital beds availability.
Meanwhile, More than 80 percent of the information that our brains are processing is visual. visuals communicate much more and most of the information, in a much faster way.
So it's important to know that visual insights of data can be more effective and also easy for people to make fi rm decisions faster.
Keeping these scientifi c considerations in mind, analysis of charts using visualization and data mining techniques we started this study. So the users will be updated to the information about each hospital and further, we can add the feature to book the beds in particular hospitals. It will be even more helpful if we could add the option to track the nearest ambulance and call concerning user location just like booking a cab. This mobile application will help patients in getting an ambulance and hospital bed with as much easy as getting as a cab.

Conclusion
In the present study, the data mining model was developed for predicting whether patients should admit or not, concerning hospitals' bed availability rate. The model developed with data mining's decision tree was found to be effi cient with the percentage of accuracy of 83.33%.
Decision tree algorithm and followed by data visualization techniques like histogram, bar graph, pie chart, and line chart were applied directly on the dataset using a python programming language.

Pie chart
A Pie Chart shows a static number and how classifi cations speak to a part of an entire structure of something. A pie diagram speaks to numbers in rates, and the all-out whole of all portions needs to rise to 100% Note: Alphabets in Pie chart are representing the Hospitals.

Line chart
A line diagram or line plot or line diagram or curve outline is a kind of graph which shows data as a progression of data points called 'markers' associated with straight-line fragments. The developed model and visual charts would be very helpful for patients and the healthcare sector to visualize the bed availability rate and to make a fi rm decision to navigate directly to the appropriate hospital.