Senior Data Scientist, Tiger Analytics (Jan’25 - Present)
- Revenue Optimizer
- Methods
- Optimized revenue generation by applying the Method of Moving Asymptotes (MMA) to operate within defined constraints.
- Tools
- R: Utilized the
nloptr
package, a powerful tool for nonlinear optimization.
- Forecasting Time Series variable
- Methods
- AutoRegressive (AR) and Vector AutoRegressive (VAR) Models.
- Tools
- Python (
pandas
, numpy
, statsmodels
)
Data Scientist, Tiger Analytics (Aug’23 - Dec’24)
- Price Optimization through Market Analysis
- Methods
- Analyzed market performance using historical Key Performance Indicators (KPIs) to generate a composite score.
- Forecasted future KPIs using Time Series Models.
- Tools
- Python (
pandas
, numpy
) for comprehensive data analysis and manipulation.
- Excel Solver for optimization of weighting factors.
- Representative Property Identification for Market
- Methods
- Performed K-Means clustering analysis on property and location attributes to group similar properties.
- Applied Local Outlier Factor (LOF) technique to identify and remove outliers within each cluster.
- Calculated cluster centroids and used cosine similarity to identify representative properties for each market.
- Tools
- R (
data.table
, Rlof
) for data analysis.
- R (
leaflet
, mapview
) for interactive geospatial visualization and map generation.
- Finding Optimal Leases for Discussion
- Methods
- Employed Fedorev’s Exchange Algorithm to generate optimal hypothetical lease scenarios.
- Utilized cosine similarity to select the most representative and relevant optimal leases.
- Tools
- R:
AlgDesign
package for implementing the Federov’s Exchange Algorithm.
- Excel: Developed a macro to automate the generation of an Excel dashboard from tabular data, facilitating clear and concise presentation of results.
Senior Analyst, Tiger Analytics (Jul’21 - Jul’23)
- Clustering of Patients
- Methods
- Polytomous Variable Latent Class Analysis (poLCA) for patient group identification.
- Decision Tree Classification (rpart) for new patient classification.
- Tools
- R (
poLCA
, rpart
, openxlsx
)
- Excel Dashboard for user-friendly results visualization.
- Warehouse Rent Prediction for different Win Probability
- Methods
- Linear Regression, Gradient Boosting Machines (GBM), XGBoost for predicting Rent.
- Logistic Regression (for win probability)
- Tools
- R (
dplyr
, gbm
, xgboost
, ggplot2
)
- Dataiku
Analyst, Tiger Analytics (Jan’21 - Jun’21)
- Statistical Test of Significance
- Methods
- Statistical t-Test, z-Test, and proportion Test.
- Tools
dplyr
, tidyverse
, spss
packages in R.
Academics
- Statistical Simulation (Bayesian Estimation, MCMC)
- Tools
- R Programming
- \LaTeX for report generation
- Computation of Descriptive Statistics for Practical
- Tools
- C Programming, Excel
- R Markdown for Report Generation.