Hello, I'm
Driven by a passion for data science and artificial intelligence, I engage in data-driven intelligence, utilizing machine learning, statistical analysis, and data visualization to enhance problem-solving. This enthusiasm fosters continuous exploration of new technologies, supporting a dynamic and adaptable approach in my work.
Get To Know More
1.5 years of hands-on project experience in Machine Learning and Data Analytics
Currently pursuing
B.Tech in AI & DS
I am a third-year B.Tech student at RAIT, D Y Patil University, specializing in Artificial Intelligence and Data Science, with a solid foundation in machine learning and currently deepening my expertise in deep learning. My academic journey has involved diverse projects where I’ve built predictive models, optimized data processing workflows, and crafted insightful visualizations, all aimed at deriving actionable insights from complex datasets. These experiences have strengthened my analytical and problem-solving skills, especially in data mining, feature engineering, and model evaluation. Now, as I advance into deep learning, I am working on neural networks and complex architectures to address sophisticated data challenges, positioning me to add value to AI and data science roles where real-world applications of data-driven insights are essential.
Explore My
Experienced
Intermediate
Intermediate
Experienced
Experienced
Intermediate
Basic
Experienced
Experienced
Experienced
Intermediate
Intermediate
Experienced
Experienced
Experienced
Experienced
Experienced
Intermediate
Intermediate
Intermediate
Browse My Recent
Developed a predictive model for evolutionary processes in biological systems using an ensemble model of Random Forest, XGBoost, SVM, KNN, and Neural Networks on genetic datasets. Conducted feature extraction, data normalization, and data encoding to enhance model input quality. Utilized GridSearchCV for hyperparameter tuning and implemented soft voting for ensemble predictions. Evaluated model performance using accuracy, precision, recall, F1 score, and ROC-AUC, with visualization through confusion matrices and ROC curves to support insights into trait evolution and model reliability.
Created a browser extension that detects fake reviews on Amazon using advanced data science techniques, including multi-model comparison to ensure optimal predictive accuracy. Conducted feature engineering with TF-IDF and sentiment analysis, followed by training and evaluation of multiple machine learning models—Random Forest, Logistic Regression, SVM, and Gradient Boosting Classifier. Selected the best-performing model based on accuracy and deployed it via a Flask API for real-time predictions. Leveraged Python, scikit-learn, and NLP to build an end-to-end solution for review validation.
Developed DNA-Sequencing-Classifier to differentiate between human, chimpanzee, and dog DNA using k-mer encoding and a Naive Bayes classifier. Conducted feature extraction with CountVectorizer and optimized the model with soft voting, normalization, and encoding methods. Achieved species classification with high accuracy, precision, recall, and F1 scores, providing reliable predictions for DNA sequence classification. Documented the project with a comprehensive README and modular code organization to facilitate future enhancements, including model scalability and hyperparameter tuning.
Developed a predictive model for sales forecasting using ARIMA on Perrin Frères' monthly champagne sales dataset (1964-1972). Conducted time series decomposition, stationarity testing, and seasonality analysis to enhance model accuracy. Utilized ADF test, differencing techniques, and ACF/PACF plots for feature selection and seasonal trend analysis. Implemented model evaluation and forecasted future sales, visualizing predictions against historical data to aid in strategic decision-making.
Developed a Custom Sentiment Analysis project utilizing DistilBERT, a transformer-based model, to classify SMS messages as spam or ham. Implemented data preprocessing techniques, including tokenization and binary encoding of labels, on the SMSSpamCollection dataset. Fine-tuned the DistilBERT model using TensorFlow, optimizing hyperparameters for enhanced performance and employing evaluation metrics such as confusion matrix, accuracy, precision, recall, and F1 score to assess model effectiveness. Demonstrated strong skills in machine learning, natural language processing (NLP), and Python programming, contributing to robust spam detection solutions.
This project involves visualization, data mining, and processing to create compelling visuals that effectively convey the science and impact of climate change. By analyzing and processing data, the platform will generate meaningful insights which will be presented through engaging visualizations, enhancing public understanding and awareness of climate change issues.
Get in Touch
