Projects
Financial Complaints Dashboard
Goal:
Develop a Dashboard having "Total complaints", "Timely Response", "In Progress," Dispute Rate" and " Resolved at No Cost" are KPIs for this dataset. Also, added relevant visualization KPIs to the Dashboard.
​
Results:
By using dashboards, you will discover the right insights in an easy way. You can apply filters yourself and consult different graphs. By immediately having the right information available, correct decisions can be made.
​
​
What I Learned :
-
Use of Hex Tile Map in Tableau.
-
Unique implementation of Dual Axis to improve visualizations.
-
Charts in Tableau.
Exploratory Data Analysis of SuperStore Sales:
Goal:
To Find :
-
What is the overall trend of the sales?
-
What are the top 10 products by sales?
-
Which is the most performing Segment?
-
What is the most preferred Ship Mode?
-
Which are the most profitable category and Sub-category?
​
Process:
Data Cleaning ( removing duplicates and null values) -> Changing the format of Data according to the need ( Change of unformatted Date to Date Format) -> Analysing the Data and finding the required output.
​
Results:
-
There are increasing trends or growth in Sales over time, there may be a seasonality to the sales for each year.
-
Overall growth in sales is observed in the Months of September, November, and December.
-
The same pattern is observed each year, however, it appears at different levels.
​
​
What I Learned:
-
Kera library( to build and train models)
-
Use of Multitable analysis , for summarising the data quickly.
-
A low performance model can be improved by increasing number of neurons , but it can improve it up to certain limit.
Sales per Month
Most Preferred Ship Mode
Most Profitable Category and Sub-Category
Most Performing Segment
Predicting Car Selling Price
Goal:
Predict car selling values based on the dataset found on Kaggle by Cardekho.com using RandomizedsearchCV Algorithm.
​
Overview:
The dataset was cleaned and a new data set was developed using dependent and independent variables.RandomizedsearchCV Algorithm was used to improve the model and find the best combination.
​
Result:
This is the WebApp that would help you to predict the cost of your owned vehicle.
​
Link for the application GitHub-Link
​
What I learned:
-
Supervised Machine Learning.
-
User of RandomizedSearchCV ( Scikit-Learn API).
-
Deployment of the application to Heroku.
Global Performance Dashboards
Goal:
Develop a Dashboard where "Sales ", "Quantity", "Average Delivery Days", and " Returned Orders" are KPIs for this dataset. Also, added relevant visualization KPIs to the Dashboard.
​
Results:
By viewing this dashboard we can easily understand the performance of the company in last 5 years in Different regions.
​
​
What I Learned:
-
Use of Different charts in Power BI.
-
Analyzing tables and relation.
-
Data Cleaning using Power Query Editor with DAX.
-
Developing an Interactive Dashboard.
Currently Working On
1. Studying the effect of social approval assets of a venture and its investors on the exit outcome of the venture.
​
Workflow :
Data Collection[LexisNexis] ( Web scraping) -> Data cleaning -> Text Analysis