Puja Thakre

• MediGraph: Disease–Drug Knowledge Graph via Wikidata

- Extracted disease–drug relationships from Wikidata using SPARQL queries and stored them in a structured CSV format.

- Performed data preprocessing to balance disease distribution and merged redundant disease descriptions for cleaner insights.

- Visualized the medical knowledge graph using NetworkX and PyVis, enabling intuitive exploration of therapeutic relationships.

• InboxSecure:-

- Developed an SMS spam detection model using NLP techniques, including text cleaning, tokenization, stopword removal, and label encoding

- Performed EDA with Pandas, Matplotlib, Seaborn, and Plotly to analyze message patterns and class distribution

- Trained and evaluated classifiers (Naive Bayes, logistic regression, SVM), achieving high accuracy and validating results with precision, recall, F1-score, and confusion matrix

• BankChurn: customer retention prediction model

- Developed a supervised machine learning model to predict bank customer churn, with preprocessing steps including imputation, feature encoding, and class balancing

- Conducted exploratory data analysis using Seaborn and Matplotlib to uncover churn patterns across demographics and service usage

- Trained and compared models (logistic regression, decision tree, random forest), evaluating performance using accuracy, precision, recall, F1-score, and ROC-AUC

• HouseWorth – predicting real estate prices in Bangalore

-Conducted exploratory data analysis on the Banglore Housing Dataset, informing subsequent modeling decisions 

-Constructed housing price model with Linear Regression, Random Forest, Decision Trees, & XG-Boost 

-Demonstrated predictive modeling methods to derive insights and enhance accuracy in housing price predictions

• MoodMap – Mapping Emotions from Voice

- Extracted meaningful audio features (MFCCs, pitch, intensity, chroma) from speech recordings to represent emotional characteristics.

- Applied PCA for dimensionality reduction and used K-Means and DBSCAN clustering to group similar emotions, evaluated using Adjusted Rand Index (ARI).

- Performed data analysis and visualization to understand feature importance and cluster quality; tested the model on unseen audio data for accuracy.

• PG Coordinator, Office of Carrier Services, IIT Delhi (April 2024 - May 2025)

-Led the PG Nucleus and facilitated recruitment drives, pitching 40+ new frms for the 2024-25 session, while scheduling interviews and other placement processes for 800+ companies and 1200+ students across 25 disciplines

• RANK HOLDER: Achieved the fourth-highest rank at RTMNU University in B.Sc.
• Sukhdevji Maharaj Saongi Memorial Scolarship: Awarded by Gold Medal and Cash Prize for securing highest marks in PHYSICS and MATHEMATICS for the degree of B.Sc from Institute of Science,Nagpur
• SHRI V.S.TAMMA Silver Medal: Awarded for obtaining highest marks in PHYSICS for the degree of B.Sc.
• AASCON Scholarship: Awarded for highest marks in B.Sc. 1st year (in PCM) and in 2nd year in college