Himadri Sekhar Samanta Resume */*

HIMADRI SEKHAR SAMANTA

Austin, TX • ad3o0m@r.postjobfree.com • 240-***-**** • linkedin.com/in/himadri-samanta-77a734182 • github.com/himadriut SKILLS & PROFICIENCIES

Programming and Scripting Languages: Python (5+ years), Pandas (3+ years), C/C++ (3+ years), MATLAB (5+ years), Numpy, SAS(1+years), Mathematica(10+ years), Data Visualization: Tableau, Matplotlib, Big Data Analytics (Hadoop, HDFS, Hive, Spark, Kafka ) Operating Systems: Linux (10+ years), Windows (10+ years), MacOS (10+ years) Database Language: SQL (1+years)

Artificial Intelligence using Deep Learning: Neural Network: ANN, RNN, CNN, Natural language processing (NLP), Computer Vision, PyTorch,Tensorflow, OpenCV

Machine Learning Algorithms: Classification, Regression models, Clustering, Decision trees, ensemble techniques, model tuning, unsupervised learning, Bagging, Random Forest, Gradient Boosting, AdaBoost, XGBoost, SVM Statistical Methods: Predictive Analysis, Hypothesis testing and confidence interval, a/b testing, Statistical Inference Analytical and Computational Tools: Statistics, Applied Mathematics, quantitative research, fluid mechanics, Solving ordinary and partial differential equations, Monte Carlo and Brownian Dynamics simulation, numerical analysis, academic research, mathematical modeling, statistical modeling, data Engineering, data mining/analysis, data modeling, statistical analysis, data cleaning, manage a team, scikit-learn, Data profiling, Exploratory data analysis, research and development, coding, data structures, life science, R&D. WORK EXPERIENCE

Modjoul Greenville, SC

Data Scientist September 2022 – April 2023

• Developed a biomechanical Machine Learning (ML) model that successfully predicted workplace injuries with 90%+ accuracy by studying company databases in Pandas to locate the most vulnerable points of production.

• Calculated the twist and bend angle from the quaternion approach in Python to observe how people bend at the waist and knees and better protect workers from strains, injuries, and exhaustion.

• Constructed a regression model to calibrate a sound meter level device that minimizes environmental noise, both to create a more peaceful workplace and to detect high-pitch noises that may cause harm to ears.

• Advanced company growth by achieving $10M Annual Recurring Revenue (ARR) from scientific advancements and tools. University of Texas at Austin Austin, TX

Research Associate February 2016 – September 2022

• Authored (single author) a paper on the role of interstitial flows in collective cancer cell invasion during cancer progression and developed stochastic partial differential equations to be solved numerically using MATLAB to understand their super-diffusive motion.

o Published the results (with 98%+ quantitative agreement) in a peer reviewed journal, “Physical Rev. Research”.

• Generated several computational and mathematical models to understand the roles of biomedical and biophysical cues in collective cell invasion, intra-tumor heterogeneity of growing tumors, and identify the subpopulations responsible for a complicated process like metastasis.

o Developed stochastic partial differential equations and solved them numerically using MATLAB and Python to analyze and visualize the data, resulting in 98%+ quantitative agreement among MSD exponents with these experiments. o Brought the idea of stochastic quantization to understand complex non-equilibrium behavior, not done in the field before.

• Theorized on the concept of heterogeneity in characterizing phase diagrams and examined the role of fluctuations, in particular charge fluctuations for IDPs, utilizing C code on Brownian dynamics simulation for equilibrium and dynamical properties of IDPs. o Sped up the computation from weeks to minutes, a paradigm shifts with implications in designing protein therapeutics to battle cancer and numerous neurodegenerative diseases (e.g., Parkinson’s and Alzheimer’s).

• Published multiple scientific papers detailing the results of prior experiments and presented this research at several national and international conferences (oral presentation), research collaboration. University of Maryland College Park, MD

Research Associate March 2011 – January 2016

• Utilized polymer physics concepts to prove that protein collapse in finite-sized proteins, encoded in the contact maps of folded state architecture, is universal.

o Obtained critical knowledge on a type of protein that constitutes about 30% of human proteins, participate in numerous biological processes, and are often associated with diseases like COVID-19. o Coded Brownian dynamics simulation for protein folding in C and analyzed and visualized the data in Python. University of Sheffield Sheffield, UK

Research Associate July 2007 – June 2010

• Charted work on multiple aspects of many-body physics, and exploited field theoretic treatments to unravel previously unknown facets of Casimir forces, with particular emphasis on the role of geometry and dielectrics. Himadri Sekhar Samanta Resume 2/2

KEY PROJECTS

Comprehensive Lists: https://eportfolio.mygreatlearning.com/himadri-samanta • https://olympus.mygreatlearning.com/eportfolio McCombs School of Business (University of Texas), AI and DEEP Learning Austin, TX Natural Language Processing: Twitter US Airline Sentiment April 2023

• Scraped Twitter data for every major US airline during the month of February 2015, built a model asking contributors to classify positive, negative, and neutral tweets, and categorized negative reasons (such as "late flight" or "rude service"). Computer Vision March 2023

Tools Used: Working with images, Computer Vision, Keras, CNN

• Trained a convolutional neural network to identify and classify the plant seedlings from 10+ different species of plant based on physical properties and scientific data.

Neural Networks February 2023

Tools Used: TensorFlow, Keras, ANN, Google Collab

• Assisted the operations team in categorizing customers that were more likely to churn, building an artificial Neural Network from scratch that determined whether a client would leave the client in the next 6 months. Great Learning Online Remote

Capstone Project May 2022

Skills and Tools Used: EDA, Data Preprocessing, Customer Profiling, Bagging Classifiers (Bagging and Random Forest), Boosting Classifier (AdaBoost, Gradient Boosting, XGBoost), Stacking Classifiers, Hyperparameter Tuning in GridSearchCV

• Predicted the chances of being positive or negative for COVID-19 and identified the factors that influence diagnosis based on laboratory tests collected from suspected cases.

• Provided recommendations to the hospital on how to better manage the admission of patients to the general ward, semi-intensive unit, and intensive care units.

• Determined that a gradient boosting model gives generalized performance with high recall score, needed for minimizing the false negative, and that this model predicts with higher precision (i.e., lower false positives) and within minutes as opposed to weeks.

• Concluded that “Patient age quantile” is the most important feature followed by Rhinovirus/Enterovirus, patient admitted to regular ward, Platelets and Leukocytes.

McCombs School of Business (University of Texas), Post Graduate Programs in Data Science/Business Analytics Austin, TX Unsupervised Learning: Trade and Ahead May 2022

Skills and Tools Used: EDA, K-means Clustering, Hierarchical Clustering, Bank Customer Churn Prediction

• Analyzed data, grouped stocks based on the attributes and categories provided, and shared insights about the characteristics of each group to assist Trade and Ahead, a financial consultancy firm, in offering their customers personalized investment strategies. EDUCATION

Indian Association for the Cultivation of Science Kolkata, India Doctor of Philosophy (Ph.D.) in Statistical Physics May 2008 Awards & Honors: Member of the CSIR Fellowship

Extracurricular Activities: Peer reviewed for the Physical Review E, Physical Review Letters, Mathematical Biosciences, and Physical Biology journals, and was an editor for the PLOS One journal. Jadavpur University Kolkata, India

Master of Science in Physics, First Class (Grade A) December 2001 Batchelor of Science in Physics, First Class (Grade A) December 1999 LICENSES & CERTIFICATIONS

Cloud Computing – McCombs School of Business (University of Texas) October 2023-Present AI and Deep Learning – McCombs School of Business (University of Texas) May 2023 Post Graduate Program in Data Science and Business Analytics -McCombs School of Business May 2022 Applied Plotting, Charting, and Data Representation in Python – Coursera (Credential ID 4XHBN3ZUP8TW) May 2019 Applied Machine Learning in Python – Coursera (Credential ID 6WAD3WP3Z5TC) May 2019 Introduction to Data Science with Python – Coursera (Credential ID FQ7F37ES2CG8) April 2019 PUBLICATIONS

For full publication list: https://scholar.google.fr/citations?user=eCdkDkQAAAAJhl=en 1."Interstitial flows regulate collective cell migration heterogeneity through adhesion". Himadri Samanta (Physical Review Research 2, 013048, 2020) 2. "Charge fluctuation effects on the shape of flexible polyampholytes with applications to Intrinsically disordered proteins". Himadri S Samanta, D. Chakraborty and D. Thirumalai, (J.Chem. Phys., 149, 163323, (2018)). 3. "Casimir force induced collapse in fluids in non-equilibrium steady state". Himadri S Samanta, Mauro L. Mugnai, T R Kirkpatrick and Dave Thirumalai, (J. Phys. Chem. Lett. 10, 2788, 2019). 4. "Cell growth rate dictates glass to liquid-like transition and tumor heterogeneity".A. N Malmi-Kakkada, Xin Li, Himadri S Samanta, Sumit Sinha, and D.Thirumalai (Phys. Rev. X, 8, 021025(2018).

5. "On the origin of super-diffusive behavior in a class of non-equilibrium systems". Himadri S Samanta, et.al. (Phys. Rev. E 99, 032401, 2019)

Contact this candidate