.
CONTACT
Address: Clarksburg MD
Phone: 202-***-****
Email: adjbjr@r.postjobfree.com
LinkedIn: Haritha Guttikonda
GitHub: Haritha1298
EDUCATION
Masters in Data Science
University of Virginia, Charlottesville VA
Bachelors in Electrical Engineering
IIT Bhubaneswar, Odisha India
CERTIFICATION
AWS Certified Machine Learning – Specialty
PUBLICATION
IEEE SIEDS 2020: S. Choudhary, H. Guttikonda,
D. R. Chowdhury and G. P. Learmonth,
"Document Retrieval Using Deep Learning,"
2020 Systems and Information Engineering
Design Symposium (SIEDS), Charlottesville, VA,
USA, 2020.
SKILLS
Languages & Databases:
Python, R, SQL, MATLAB, JAVA.
Tools & Frameworks:
Pyspark, TensorFlow, PyTorch, Keras, XGBoost,
AutoML, NLTK, ElasticSearch, Tidyverse, dplyr,
Carat, GGPlot, Plotly, Matplotlib, Qlik Sense,
Power BI, Tableau, AWS.
Data Scientist
Proficient data scientist with demonstrated ability in synthesizing and communicating high-impact insights and recommendations to business and research questions and concerns. Key strengths include data cleaning, data engineering, modeling, statistical analysis, and creative problem-solving skills. PROFESSIONAL EXPERIENCE
Microbiome Data Scientist, 07/2020 to Current
Trans University Microbiome Initiative, University of Virginia
• Working on predictive models to understand the influence of microbiome human health and disease.
• Develop computational data pipelines using Nextflow for processing and analysis of next generation sequencing data to automate redundant processes.
• Built and deployed high-throughput bioinformatics data products with web interface to visualize data and experimental outcomes. Research Assistant, 01/2020 to 07/2020
School of Medicine, University of Virginia
• Developed a predictive model in Python to flag infants susceptible for sepsis based on 800k rows of vital signs data accounting for 726 infants over a period of study with only 106 events of sepsis.
• Proposed Survival Analysis to model for 'time-to-event' metric which increased the scope of medical research by providing hazard function for each patient.
Data Science Capstone Intern, 09/2019 to 04/2020
Logistic Management Institute, LMI
• Modeled to classify and retrieve 2000 unstructured text documents using BERT and HAN with 98% and 95% accuracy, respectively.
• Extracted output of intermediate layer in the keras model to get vector representations for the documents which is used to cluster the documents and achieved an adjusted rand index of 0.98.
• Achieved 0.85 MAP score with 10% improvement over baseline TF-IDF model by employing ensemble of TF-IDF and BERT models. ACADEMIC PROJECTS
DeepFake Image Detection
• Trained classification model using DenseNet and VGGFace frameworks to detect fake images generated using GANS; achieved 97% accuracy over 140,000 images.
Bayesian Inference for Wildfire Crisis Management using Tweets
• Implemented Bayesian Belief Network and Naive Bayes Classifier on wildfire tweets dataset to determine the relevancy and importance of the tweet. Predictive Analysis for Home Automation using ARIMA & RShiny
• Predicted solar irradiance and grid price using ARIMA to reschedule non- critical loads to non-peak hours dynamically to optimize solar energy consumption and electricity bill.
.
HARITHA
GUTTIKONDA