In this repository, we go over some of the visualizations I've created personally and by participating in London Business Analytics Group Projects using PowerBI.
Below are some of the project previews. Click on the links to view the full description of each project
We explore a famous machine learning dataset, the Breast Cancer Wisconsin (Diagnostic) data, to see if it is possible to build a model with some predictive power. We visualise the data as histogram, box and whisker plot and a ‘home-made’ swarm plot. To do this, we need to shape the dataset in the Query Editor and standardise the data with a few DAX calculations. Below are some of the visuals
Using PowerBI, I visualize and analyse the Telecommunication dataset to gain insight into customer attrition and make business decisions. I created few visuals to understand why customers are churning the comapny. Using Key Influencer visual, I was able to gain insight into the key attributes that makes customers churn the comapny and also the combined attributes impact on customer attrition using Top segment of the Key Influencer visual. These attributes are later used for prediction with Python.
Using PowerBI, I clean up a famous dataset - a passenger list from the Titanic. Using Power BI's Query editor, I rename some columns, split others apart, add new columns and replace obscure values with more meaningful ones. I then create a few visuals to understand who survived and why - was it women and children first or were other factors important? I build ribbon charts to show the age profile and use the Key Influencer visual. I employed interactive techniques such as cross-highlighting and drill-through pages to get more insight.
English Premier League(EPL) is my favorite football league and I am a supporter of Manchester United. Using dataset from EPL websites, I build an EPL interactive league table with PowerBI. To achieve this, I used the Query Editor to split, reshape, sort, append, calculate on then group the data. I then built the table visualisation, then improve it with conditional formatting and custom tooltips to provide additional information of each team when one hover on it.
