Problem Statement:
-
Create a storyboard showing spread of Covid-19 cases in your country or any region (Asia, Europe, BRICS etc) using Tableau, Power BI or SAP
-
Identify interesting patterns and possible reasons helping Covid-19 spread with basic as well as advanced charts
Abstract:This notebook is devoted to predicting the COVID-19 spread dynamics in the world using neural networks of different structures. We will see how to make predictions based on linear regression, back propagation, long short-term memory neural networks(LSTM) , autoregressive integrated moving average(ARIMA) and prophet.
Steps or Stages: It consists of four stages:
-
Download and preliminary analysis of data
-
Linear Regression
-
Back Propagation NN
-
Long Short-Term Memory - LSTM
In first stage we pre-prepare data for analysis: -change the data types of columns
-
Rows Filtering
-
Elimination of Missing Data
-
DataSet Transformation
-
Data Normalization
Prerequisites:
-
Python
-
Pandas
-
Statistics
-
NumPy
-
Matplotlib
-
Keras
-
Scikit-Learn
Dataset:https://www.kaggle.com/shiv28/covidowid
About: We are doing forecasting on India and taking train:test = 7:3.
Results:
Linear Model:
ANN:
LSTM:
Compare between Linear Model , ANN , LSTM
ARIMA:
Prophet:
