Komal Maniar

Explored the Citibikes trips & stations dataset and revenue dataset for New York City
Implemented Marketing Analytics techniques like Forecasting and Seasonality etc using SAS to get key insights on the revenue dataset
Data Visualization using Tableau to study demographics of the type of users, their gender, age group, most popular and least popular stations etc. to optimize the efficiency of the Citibikes Stations Network
Devised marketing strategies based on the results of the… Show more

Explored the Citibikes trips & stations dataset and revenue dataset for New York City
Implemented Marketing Analytics techniques like Forecasting and Seasonality etc using SAS to get key insights on the revenue dataset
Data Visualization using Tableau to study demographics of the type of users, their gender, age group, most popular and least popular stations etc. to optimize the efficiency of the Citibikes Stations Network
Devised marketing strategies based on the results of the analyses for better revenue generation

Show less

• Built quantitative models and learnt the execution of Bootstrapping, Monte-Carlo Simulation etc. quantitative techniques in Python using NumPy, Pandas, SciPy libraries

Designed & integrated an enterprise architecture for SOS: helps in reducing the response time for 911 services in case of emergencies
This technology can help victims even when they are not in a situation to actually place a call to 911 for help
Prepared a comprehensive business case comprising of Scenario Analysis, identification of KPIs, market research, strategic goals, cost-benefit analysis, success metrics (session length and API latency) and SWOT analysis

Twitter Sentimental Analysis on the tweets sent out by active users, to understand their opinions as positive, negative or neutral after the First GOP Presidential Debate in the United States 2016
The sentimental analysis was done based on the Candidates, the subject of the tweets etc. using R
Constructed Network Graphs with candidates as nodes and the edges connected only if they are mentioned in the same tweet.

Designed & integrated an enterprise architecture for Let’s; enabling students to reconvene and connect over common interests like academics, arts and sports
Prepared a comprehensive business case comprising of Scenario Analysis, identification of KPIs, market research, strategic goals, cost-benefit analysis, success metrics (session length and API latency) and SWOT analysis

Developed a statistical model analyzing factors influencing voting patterns for presidential candidates at Primaries using SAS
Performed Exploratory Data Analysis to identify collinearity outliers in demographic data spanning across factors ranging from household income, literacy rate and spend trends
Implemented Principal Component Analyses and regression to evaluate independent variables peculiar to voter inclinations based on their correlation with the nominees’/candidates… Show more

Developed a statistical model analyzing factors influencing voting patterns for presidential candidates at Primaries using SAS
Performed Exploratory Data Analysis to identify collinearity outliers in demographic data spanning across factors ranging from household income, literacy rate and spend trends
Implemented Principal Component Analyses and regression to evaluate independent variables peculiar to voter inclinations based on their correlation with the nominees’/candidates
Validated the models, running them on test dataset that reflected an 85% accuracy
Show less

Constructed an ER diagram for college administrative purpose to enhance the efficiency of data retrieval
Executed complex nested MySQL queries using joins, groupings and aggregation for data manipulation and analysis

An Implementation and Comparative study for Principal Component Analysis (PCA) and Multilinear Principal Component Analysis(MPCA) for a Face Detection and Recognition System.

Evaluated the techno-commercial feasibility of Light Motor Vehicle based on price, safety, maintenance and design using R-Rattle GUI
Employed a CRISP-DM methodology - C4.5 algorithm classifying target variable, thereby forming decision rules