Developed and deployed a robust machine learning solution for Human Activity Recognition (HAR) using sensor data. Utilized a diverse set of algorithms including Support Vector Machine (SVM), Random Forest, and Gradient Boosting, achieving high accuracy and robust generalization. Conducted extensive Exploratory Data Analysis (EDA) and recommended model selection strategies for real-world applications

Designed and executed a sentiment analysis system, integrating NLP techniques and machine learning models. Proficient in data preprocessing, class imbalance mitigation, and model selection, yielding a 77% accuracy rate. Explored diverse model architectures, including Naive Bayes, Logistic Regression, LSTM, and GRU, and leveraged pre-trained word embeddings and spaCy for advanced NLP capabilities.

Collaborated on a project as my bechelor’s project to develop a Urdu character recognition system. Engineered synthetic datasets, explored various machine learning algorithms (KNN, SVM, MLP, Deep Learning, Random Forest), and conducted comparative analysis to enhance Urdu character recognition.

Developed a clinic management system using Python and Django framework. It allowed patients to book an appointment with a doctor. The system had the patient’s history for the doctor to access. Patients can access the portal to see their medications and medical history. I integrated SQLite database for data storage and builtin user authentication.

Developed a weather data analysis project using Python and SQL, involving database creation, data cleaning, visualization, and machine learning techniques. Conducted in-depth exploration of weather patterns, correlations, and network analysis to gain insights into temperature variations across multiple cities. Utilized statistical models for temperature prediction and demonstrated strong problem-solving skills and data-driven decision-making.

Developed a Python-based image processing project that utilized a series of filters and image conversions to
generate a cartoon-style output from a given image. The project involved implementing various image
processing techniques, such as blurring, edge detection, and color quantization, to achieve the desired effect.

Implemented a deep learning model, utilizing a Multi-Layer Perceptron (MLP) architecture, for binary classification in a credit risk analysis project. Conducted data preprocessing, feature engineering, and hyperparameter tuning to enhance model performance. Achieved an accuracy of up to 77.63% and successfully addressed issues of class imbalance, demonstrating proficiency in machine learning and data science.

Developed a portal for accommodation assistance using Python and Django. It allowed the user to apply for accommodation assistance. Admin would then either approve or reject the application. It was a sample prototype of the project.