DEBRISHI BOSE

*** ******* ******, ***********, ** 06226 C: 209-***-**** ac30jr@r.postjobfree.com https://www.linkedin.com/in/debrishibose/

Strong Quantitative, Analytical, Statistical skills applied within Government, Healthcare & Utility industries Hands-on experience with a variety of data sets- Longitudinal, Spatial, Biomedical, Multilevel etc. Proficient in a variety of Statistical Modeling Techniques, including familiarity with data visualization tools 3 years of supervisory experience at the Quantitative Learning Center- University of Connecticut Predictive, Mixed Modeling

Machine Learning, Classification

Bayesian Analysis, G.L.M

Categorical Data Analysis

CART, Logistic Regression

Simulations, sample size & Power

SAS, R, SPSS, M-PLUS

C, SQL, EXCEL

Psychometrics Assistant, 05/2016 to Current

CONNECTICUT STATE DEPT. OF EDUCATION – Hartford, CT Gifted & Talented students: Identification

Thorough assessment of criteria for understanding the current practices in identification of Gifted & Talented students across US

Strategical modification of current practices to improve the test scores of Gifted &Talented students Gifted & Talented students: Estimation

Unavailability of estimates of the number of Gifted & Talented students in some public schools in US due to nonexistence of Gifted & Talented programs in such schools Data was split into two components based on absence or presence of Gifted Programs Model developed using the first part of the data set was utilized in predicting the number of Gifted & Talented students where such programs did not exist

Sample size requirements : 3-level Organizational models Prior research focused only on two level models, even though three level models are quite common in educational data

A Random-Intercept model was utilized to conduct the simulation study using 27 simulating conditions Issues involving model convergence, estimation of fixed effects, variance components, and their associated standard errors were analyzed

Comparison of Multi-Stage Testing & Computer Adaptive Testing Conducting a thorough literature review on comparison of performances of Multi-stage Testing and Computer Adaptive Testing based on matching Psychometric properties PROFESSIONAL SUMMARY

RELEVANT SKILLS

PROFESSIONAL EXPERIENCE

Identifying advantages of one over the other under various constraints, and thereby suggesting potential testing mode modifications

Statistician, 05/2013 to 07/2013

NORTH EAST UTILITIES – Mansfield, CT

Developed a system (model) to predict damage to power lines using data on weather forecasts. Achieved in three steps- classifying regions according to the degree of damage (minimal/moderate/severe) using Heat-Maps in R, identifying important predictors through Variable Selection Techniques and finally fitting a Poisson Regression model to the data.

Predictive Modeler, 01/2013 to 05/2013

Cigna HealthCare – Bloomfield, CT

Developed an unified measure of the construct: Socio-Economic Status (S.E.S) Utilized Factor Analysis via SAS-EG to identify important components of SES, accounting for at least 75% of its variability

Graduate Research Assistant, 01/2012 to 05/2012

University Of Pune – Pune, MH

Master's Thesis: Improvement of Classification Models Objective of the project was to increase the accuracy of prediction of Cancer Employed Bagging and Boosting to improve the performance of Decision Trees Comparisons made to check the percentage of improvement using the two techniques Master of Science: Educational Statistics and Research Methods, 2017 University of Connecticut - Mansfield, CT

Master of Science: Statistics, 2014

University of Connecticut - Mansfield, CT

Northeastern Educational Research Association Conference (NERA) Travel Award Recipient, Oct 2017, University of Connecticut

International Indian Statistical Association Travel Grant Recipient, December, 2012, University of Connecticut

Pfizer Scholarship Recipient: Outstanding Student Award, Fall 2010- Spring 2012, University of Pune, India

Entrance Exam (MS Statistics- University of Pune): All India Rank:4, State Rank:1 Effect of small sample sizes on parameter estimation in a three-level organizational framework (NERA Conference, October 2017. Trumbull, CT)

Development & Validation of an Instrument to measure Attitude of undergraduate students towards Statistics (NERA Conference, October 2017. Trumbull, CT) EDUCATION

ACCOMPLISHMENTS

PUBLICATIONS/CONFERENCE PRESENTATIONS

Contact this candidate