• Performed text analysis using R for twitter data set for 10 years, calculating word frequencies and analyzing the changes in twitter trend over the years. • Demonstrated and presented the insights using bigram networks, histograms, word clouds and using Zipf’s law.

• Interpreted the Keyword Co-occurrence (KCN) network between 248 keywords occurring in a list of 49 items using network graphs. • The network analysis facilitated the understanding of the relationship between the keyword pairs using network plots and scatter plots between degree and strength.

• Applied unsupervised clustering algorithms (K-means and Hierarchical) to obtain a homogeneous classification of different countries based on various attributes such as GDP, Mortality Rate, Life Expectancy, etc. • Visualized the clustering results using dendrograms and scatter plots & obtained internal and external validation measures for evaluating the performance of clustering quality.

This project explores the bike sharing data and tries to extract KPIs that might be useful in increasing the sales of rental bikes. These indicators highlight the most favorable season, temperature, weather conditions etc., for rental bikes usage. It also shows the most frequently used routes and the scarcely used routes.