DINESHKUMAR KUMAR

*****, ****** ***** **, *****, FL 33613 Phone: 813-***-**** Email: ac4913@r.postjobfree.com

PROFESSIONAL SUMMARY

Data analyst and Modeling entry level specialist with cooperate and educational experience in Data analysis, Database management, Data visualization and Client-side Scripting. A Machine learning developer focused on developing and implementing Machine learning models that facilitates reliable decision making. A dynamic thinker with ability to develop effective strategies to troubleshoot various technical issues.

AREAS OF EXPERTISE

• RDBMS • Data analytics • Machine Learning • ETL • JavaScript • Linux

• Software Development • Software Testing

TECHNICAL SKILLS

Operating System: Windows and Linux

Languages: C, C++, Java, R, SQL, Python, JavaScript, HTML

Technologies: MS SQL, PostgreSQL, Tableau, Apache Hadoop

PROFESSIONAL EXPERIENCE

Verum Properties LLC - Data Analyst and Modeling Intern Jan 2018 – Present

Developed an automated valuation model to compute capitalization rate for NNN retail property.

Delivered real time Cap rate estimates, benefiting investors by saving valuable time & resource.

Implemented time series model to forecast future asset value by intercepting trends & patterns.

Effectively managed company’s Real estate historical data via local hosting and backup.

Ensured model accuracy by real time update of training data and valuation model

Achieved faster evaluation of real time cap rate via Web - Database integration

Reduced error rate from 9.35 % to 8.65% by identifying insignificant variables in the model

EDUCATION

University of South Florida, Tampa - FL August 2016 – May 2018

Master of Science, Computer Science GPA: 3.24/4.0

Pondicherry Engineering College, India August 2012 – July 2016

Bachelor of Technology, Computer Science Engineering GPA: 7.76/10

ACADEMIC PROJECTS

NBA Analytics project

Constructed a predictive model to forecast draft value of NBA players

Identified trends and patterns that effects player’s salary, via visualization

Improved model accuracy to 75.04% via hyperparameter Tuning

Performed Ad Hoc analysis to identify various stats using database querying.

Recommendation Systems using Collaborative Filtering

Designed a movie recommendation model offering user specialized movie suggestion.

Achieved an accuracy of 63.2 % for prediction and a precision of 0 .23 for Recommendation

Increased user engagement by 2%, compared with Kaggle competition model &dataset

Contact this candidate