Novel Corona Virus : Exploratory Data Analysis along with prediction by using Machine Learning Algorithms.

A new infection based on SARS Corona virus is indeed deadly originated in Wuhan, China and spreading to other parts of world.

Background

2019-nCoV is a beta corona virus, like MERS and SARs, all of which have their origins in bats.

This dataset has daily level information on the number of affected cases, deaths and recovery from 2019 novel coronavirus.

The data is available from 22 Jan 2020.

Data

From WHO— On 31 December 2019, WHO was alerted to several cases of pneumonia in Wuhan City, Hubei Province of China. The virus did not match any other known virus. This raised concern because when a virus is new, we do not know how it affects people.

So daily level information on the affected people can give some interesting insights when it is made available to the broader data science community.

The data has daily level information on the number of affected cases, deaths and recovery from 2019 novel coronavirus.

The data is available from 22 Jan 2020.

John Hopkins Dashboard is the source of Data: https://docs.google.com/spreadsheets/d/1yZv9w9zRKwrGTaR-YzmAqMefw4wMlaXocejdxZaTs6w/htmlview?usp=sharing&sle=true#

Reading The data

Last Update, Confirmed, Deaths and Recovered are important columns

Visualizing

Deaths toll seems 100% in China

Countries having some cases of Virus →China with MAX cases

Cases Confirmed in China States

Case Fatality Rate In Chinese States

Calculating and plotting Local Outlier factor to check whether anomaly in data-set

LOF Score

K-means Clustering

Cluster formation

Applying Various Models to data-set

Linear Regression

Actual and predicted values

Simple Visualization

Applying Neural Network Model by using Stochastic gradient descent optimizer

Model Accuracy seems to be 80%

Trying one more model using “adam” optimizer and “sigmoid” activation function

Model accuracy is still 80% no change of optimizer.

Conclusion:- With more data-set i.e. real-time coronavirus can be predicted state wise in terms of spread before area being affected.

References:-

1. https://www.cdc.gov/coronavirus/2019-ncov/summary.html
2. Optimizers:- https://keras.io/optimizers/
3. Information regarding Coronavirus:- https://hub.jhu.edu/novel-coronavirus-information/