Project 1 - EDA and visualization on the Annual Vital Statistics Report-CRS for the years 2011 through 2016.

Report Link : http://www.ejanma.karnataka.gov.in/frmVitalSat.aspx

• Packages Used : tabulizer,reshape,ggplot2
• Key Tasks
  • Extracted data accurately from the given PDF files done semi-automatically, i.e., without having to re-type the data.
  • Load the data into Rstudio
  • Compute basic statistics for the data using R (i.e., min, max, mean, median, mode, variance, std deviation, IQR, etc.)
  • Detected any outliers in the chosen data
  • Produced different plots using R - simple scatter plots, bar graphs, line graphs, histograms etc.

Project 2 - IPL 2019 Data Analysis.

Dataset Link : https://www.kaggle.com/nowke9/ipldata and also extracted batsman and baller rankings from cricbuzz

• Packages Used : tabulizer, dplyr, ggplot2, reshape2, magittr, tidyr
• Key Tasks
  • Extracted data for individual player with corresponding run in each match.
  • Descriptive statistics and coefficient of variance of top 10 players.
  • Descriptive and inferential statistics of IPL 2019 and plots.

Project 3 - Time Series Analysis of UK Driver Death Dataset

Dataset - inbuilt timeseries dataset of uk driver death in r

• Packages Used : ggplot2, Metrics, forecast, reshape
• Key Tasks
  • Built a timeSeries object with the data.
  • Ploted the yearly mean values.
  • Decomposed the time series using stl function.
  • Obtain residue after removing trend and seasonality.
  • Built a model using HoltWinters model for the period upto about 75% of the data.
  • Predicted the values for the next 25% of the time.
  • Built an ARIMA model for the period up to about 75% of the data.
  • Plotted time series plots.
  • Found out ARIMA works better than Holtwinters.

Project 4 - Text Analytics on Demonitization Twitter Data

Dataset - https://www.kaggle.com/arathee2/demonetization-in-india-twitter-data

• Packages Used : ggplot2, readr, tm, wordcloud, plyr, lubridate, syuzhet
• Key Tasks
  • Preprocessed 15000 tweets to tokenize, lemmitize, count word frequencies.
  • Find top common most occuring words
  • Performed Sentimental Analysis on tweet data
  • Created cluster into group of related messsages
  • Created word cloud and other visualizations
  • Conducted test of hypothesis