Last Mile Delivery Optimization
A feasibility study to incorporate efficient routing optimization to provide delivery services in Tempe, AZ.
• Generated mathematical model in AMPL using CPLEX. Solved traveling salesman problem (TSP) in AMPL for intercity distances of 15 cities avoiding sub-tours and implementing the cut-set algorithm.
• Utilized TSP Branch and Bound method to obtain the optimum solution using assignment constraint and conducted dynamic programming for sub-grouped cities.
• Utilized TSP Branch and Bound method to obtain the optimum solution using assignment constraint and conducted dynamic programming for sub-grouped cities.
Decision Support System for Airport Authority
Reducing latency in decision making and setting up database to support airport authority.
• Analyzed multiple airlines, aircraft working at international and domestic routes and generated a SQL database.
• Used SSMS and Visual Basics for Application to make windows based software for maintaining the database.
• Used SSMS and Visual Basics for Application to make windows based software for maintaining the database.
Crafting Soccer's Chaos into Statistical Structures
Implementing Regression Analysis using Soccer Statistics.
• Generated a model that can predict the rating of any soccer team using statistics available for any match.
• Model analysis performed; Tests for model adequacy, multicollinearity, transformation and model validation.
• Generated a model that can predict the rating of any soccer team using statistics available for any match.
• Model analysis performed; Tests for model adequacy, multicollinearity, transformation and model validation.
Time Series Analysis in Healthcare Management
Waiting times for treatment are key health-policy concern in several countries. The amount of time a patient waits to be seen is one factor which affects utilization of healthcare services.
• Performed Time Series Analysis on the total number of people in waiting list for treatment at National Health Services in London from 2000 to 2008. Gathered potential models from JMP for first 80% of the dataset based on significance analysis and then performed cross-validation with remaining 20% of the dataset to select the best model with best forecast capability.
• Examined Exponential Smoothing methods, ARIMA models with seasonal effect and Vector Auto Regression models for this project.
• Performed Time Series Analysis on the total number of people in waiting list for treatment at National Health Services in London from 2000 to 2008. Gathered potential models from JMP for first 80% of the dataset based on significance analysis and then performed cross-validation with remaining 20% of the dataset to select the best model with best forecast capability.
• Examined Exponential Smoothing methods, ARIMA models with seasonal effect and Vector Auto Regression models for this project.
