Rahul Patil

Relational Database model for Online movie ticket booking system- JusTicket

• Created an Entity-Relationship diagram and Relational Database Schema of the database.
• Implemented the database in Microsoft Access with MS Access front-end
• Created data input screen forms for tables/views in Access.
• Inserted synthetic data using the input screen forms
• Designed report forms based on interesting queries.
• Building a menu-driven environment for the database system using… Show more

Relational Database model for Online movie ticket booking system- JusTicket

• Created an Entity-Relationship diagram and Relational Database Schema of the database.
• Implemented the database in Microsoft Access with MS Access front-end
• Created data input screen forms for tables/views in Access.
• Inserted synthetic data using the input screen forms
• Designed report forms based on interesting queries.
• Building a menu-driven environment for the database system using Switchboard

Show less

Through extensive Exploratory Data Analysis & innovative logical approaches, we identified possible fraudulent claims from multiple linked Health Insurance Claim data sets. In the process, we also developed Business Rules for prevention of the same in the future, to minimize loss of revenue.
We made use of SAS Enterprise Miner and Microsoft Excel for the same.

• Created the data connection, loaded the data and joined tables
• Identified dimension and measures
• Used different aggregation levels to get appropriate insights
• Developed dashboard to identify countries having highest prices and the price fluctuation
• Created a story to showcase the variation of food prices

• Built a prediction model using C5.0 algorithm which was 87.94 % accurate and identified the most important variables
• Cleaned the data, handled missing values and tackled the class imbalance problem using SMOTE function

Multiple Linear regression analysis and Principal component regression analysis on General Motors collected data on U.S. Standard Metropolitan Statistical Areas (SMSA’s) on air pollution contribution to mortality

Multiple Linear regression analysis:

• Performed Multiple Linear regression on original dataset keeping Mortality as Target variable
• Analysed significant variables using hypothesis analysis where P-values are less than 0.05 or 5% and again performed regression… Show more

Multiple Linear regression analysis and Principal component regression analysis on General Motors collected data on U.S. Standard Metropolitan Statistical Areas (SMSA’s) on air pollution contribution to mortality

Multiple Linear regression analysis:

• Performed Multiple Linear regression on original dataset keeping Mortality as Target variable
• Analysed significant variables using hypothesis analysis where P-values are less than 0.05 or 5% and again performed regression analysis keeping significant variables and investigated the regression summary and Fitness of the data model looking at a coefficient of Correlation and Determination and standard error.

Principal component regression analysis:

• Executed Principal Component Analysis on original dataset keeping Mortality as Target variable and evaluated 5 principal components
• Performed Multiple Linear regression on selected 5 principal components
• Analysed significant variables using hypothesis analysis where P-values are less than 0.05 or 5% and again performed regression analysis keeping significant variables and investigated the regression

Conclusion :

First regression model fails to determine whether air pollution is significantly related to mortality, because of HCPot, NocPot and SO2Pot are not significant as p-value is greater than 0.05 (or 5%), instead, it indicates the positive significance of %Non-White to Mortality.
However, the principal component regression analysis strongly indicates air pollution is significantly related to mortality. The PC1, which has the highest Eigenvalue (4.0289) and contributes to a maximum proportion of 28.78%, is strongly correlated with the variables recording the pollution potential of three different air pollutants (HCPot, NocPot, and SO2Pot).

Hence it determines air pollution is significantly related to mortality and we would prefer regression model obtained by considering significant principal components.
Show less

This dataset contains house sales prices for king county, the USA which includes Seattle. It includes houses sold between May 2014 and May 2015. This dataset enables us to study and evaluate simple regression models. There are 21 variables in the dataset and 29000 records in the dataset.

dataset Link : https://www.kaggle.com/harlfoxem/housesalesprediction

• Performed statistical exploration using Excel and SAS E-Miner on raw datasets and conducted trend analysis
• Analyzed… Show more

This dataset contains house sales prices for king county, the USA which includes Seattle. It includes houses sold between May 2014 and May 2015. This dataset enables us to study and evaluate simple regression models. There are 21 variables in the dataset and 29000 records in the dataset.

dataset Link : https://www.kaggle.com/harlfoxem/housesalesprediction

• Performed statistical exploration using Excel and SAS E-Miner on raw datasets and conducted trend analysis
• Analyzed King County housing dataset by generating Principle component analysis and correlation matrix
• Predicted housing prices on King County data multiple linear regression analysis
Show less