Sharmilashree Vanniaperumal

Enthusiastic and detail-oriented Health Care Data Analyst with a strong passion for leveraging data to improve healthcare outcomes. Experienced in working on projects involving machine learning and natural language processing to analyze and interpret complex healthcare data.

Predictive Modeling for Autism Spectrum Disorder  Sep2023-Oct2023

• Developed predictive models using Logistic Regression, SVM, KNN, and Random Forest algorithms to predict autism spectrum disorder.
• Achieved a predictive accuracy of 99%.
• Technologies: Google Colab, Jupyter, MATLAB, Seaborn, TensorFlow, Python.

Homicide Data Analysis  Oct2023-Nov2023

• Conducted an in-depth analysis of homicide data and developed a Decision Tree Classifier.
• Achieved a predictive accuracy of 96%.
• Technologies: Google Colab, Jupyter, MATLAB, Seaborn, TensorFlow, Python.

Machine Learning Project: Predicting Diabetes Status  Nov2023-Dec2023  

• Analyzed lab results from Iraqi patients using 12 machine learning models.
• Top performers were Random Forest, XGBoost, and Bagging Classifier, achieving 99% accuracy.
• Technologies: Google Colab, Jupyter, MATLAB, Seaborn, TensorFlow, Python.

Predicting Significant Variables for Diabetes Nov2023-Dec2023

• Employed statistical methods and non-parametric testing to identify influential factors for diabetes.
• Technologies: RStudio, Data Visualization.

BERT-based Text Classification Project     Jan2024-March2024

• Implemented a BERT-based classifier for sentence classification on the IMDb dataset.
• Fine-tuned the model to achieve high performance metrics in sentiment analysis.
• Technologies: Python, PyTorch, Hugging Face’s Transformers Library, BERT.

Predicting Drug-Target Interactions Using Language Models  Jan2024-May2024

• Developed a Siamese Neural Network model using SMILES and amino acid sequences to predict drug-target interactions.
• Achieved an AUROC of 0.8777 and an AUPRC of 0.6922.
• Technologies: Python, PyTorch, ChemBERTa, ProtBERT, Neural Networks.NLP

Predictive Model for Alzheimer’s Disease         Jan2024-Feb2024

• Created machine learning models to classify stages of Alzheimer’s disease using demographic, clinical, and neuroimaging data.
• Achieved accuracies up to 91.70% with models like Random Forests.
• Technologies: Python, Scikit-Learn, Pandas, NumPy, Matplotlib, Seaborn.

Text Classifier Development for NLP        Jan2024-Feb2024

• Developed a neural network-based text classifier using a Deep Averaging Network (DAN).
• Achieved a benchmark accuracy of 92.24% on development data.
• Technologies: Python, PyTorch, NumPy.