aryan1113/forage-data-visualization

Files for courses I'm pursuing from Forage (a platform by TATAIQ )

Forage-data-visualization

Files for a courses on Forage (a platform by TATAIQ )

The ANZ Australia Data Analysis has a sysnthesized transactional-dataset of 100customers, which represent real customers. Descriptive analysis of the dataset is done.

• Missing value treatment.
• Outlier Detection and Treatment.
• Basic Visualization and Plotting of various columns

TATA Data Visualization

Created visualizations in Tableau

• Line-plot to see (trendline) variation of data with time.
• Visualizing top-10 countries based on revenue.
• Finding out most valuable customer.
• Using a map-plot.

forage-data-visualization's Issues

Replace datapoint values with median, whose z-score is >=3

For the ANZ Australia Data Analytics and Visualisation dataset =>

For columns amount and balance, we have huge number of outliers (15% and 10% respectively), so you cannot drop the rows.
Instead, I've tried to log-normalize the data, this reduces the outlier % drastically (from 15% to 8.6% for amount column and from 10% to 4.3% for balance column.

This still is a high outlier %, to treat this, replace outlier data-points with median value (as these outliers have huge impact on the mean)

Add city column to dataset

For the ANZ Australia Data Analytics and Visualisation dataset =>

Using Longitude and Latitude column created, use any method to get the city name and store it in a new column.
Further data-visualization can be performed using this new column :)