Hey 👋, I'm Rajan Gupta

Data Collection and Storage: SQL, MySQL and Postgres.
Data Processing and Analytics: Python, Jupyter and Spyder.
Development: Git, Linux and Docker.
Data Visualization: Seaborn, Matplotlib and Tableau.
Machine Learning: Classification, Regression, Clustering, Decision Trees, Ensemble, LDA, CNN, Naive Bayes, Bagging and Boosting
Machine Learning Deployment: Flask and Heroku.

Programming Languages: C/C++ ,JavaScript ,SQL ,Python
Python library: PySpark ,Keras ,Python (eg. scikit-learn ,numpy ,pandas ,matplotlib), Data Visualization(eg. Plotly ,Matplotlib ,seaborn) ,PyTorch.
Technologies/Frameworks/Others: ETL ,Data science pipeline (cleansing, wrangling,visualization ,modeling ,interpretation) ,Statistics ,Time series, BootStrap ,Django ,OOPS ,TensorFlow ,Amazon Web Services (AWS) ,streamlit ,Flask ,WebScraping ,Jupyter Notebook ,APIs ,Excel ,Git ,BLOCKCHAIN.

Data Science Projects

Heart Disease Prediction (Machine learning)—Python, Flask, LogisticRegression|
Developed Advanced Machine Learning Model: Created a sophisticated machine learning model to accurately predict the likelihood of heart disease based on patient data, showcasing expertise in predictive analytics, healthcare applications, and advanced modeling techniques
Conducted feature engineering and performed exploratory data analysis to identify key risk factors contributing to heart disease. Through feature importance analysis, successfully identified top predictors and achieved an F1 score of 0.82, demonstrating the model’s ability to accurately classify patients’ heart disease status.
Dog breed prediction(Deep learning)—Python, ML, Flask, TensorFlow-Hub, Keras, Scikit-learn)

Dog Breed Classification Web Application: Developed a web application using Flask, allowing users to upload images and predict the breed of a dog using a deep learning model trained on TensorFlow 2.3.0 object classification. The model, based on transfer learning and Inception v3 as the base model for feature extraction, achieved high training accuracy of 99% and testing accuracy of 94% on a dataset of 30 famous dog breeds.
Convolutional Neural Network (CNN) for Dog Breed Prediction: Utilized Keras to build, train, and test a convolutional neural network, demonstrating proficiency in deep learning techniques. The model employed transfer learning, leveraging the Inception v3 architecture, and was trained on approximately 2.5k images of 30 dog breeds. The deployed model, with Flask as the backend, provided accurate dog breed predictions based on user-uploaded images.