I'm a Bachelor's student in Computer Science with a focus on Artificial Intelligence. I am passionate about Computer Vision, Natural Language Processing and Supervised Learning applications, and I am eager to apply my theoretical knowledge to practical experiences.
- Languages: Python, JavaScript, R
- Data Science: Pandas, Numpy, Matplotlib, Altair, Seaborn, Tableau, Jupyter Notebooks
- Machine Learning & AI: PyTorch, TensorFlow, Keras, Scikit-Learn, Optuna, OpenCV, NLTK
- Web Development: HTML5, CSS3, React.js, Django (Backend)
- Database Management: SQL, MongoDB
- Computer Vision:
- Rocket Objects Detection: Developed a robust YOLOv8m-based model to accurately identify and localize various rocket components. Achieved an overall precision of 0.835 and mAP50 of 0.826. Demonstrated the model's capability on a real-world video: ULA Atlas V launch.
- Recyclable and Household Waste Image Classification: Developed and fine-tuned a deep learning model based on the EfficientNet B3 architecture for accurate classification of waste types from images. Achieved a test accuracy of 93.54%.
- Butterfly Species Image Classification: Developed and fine-tuned a deep learning model based on the ResNet50 architecture for accurate classification of butterfly species from images. Achieved a test accuracy of 94.62%.
- Data Science & Supervised Learning:
- Predictive Modeling for Diabetes: Developed and evaluated various ML models to predict diabetes based on medical and demographic features. Achieved 90% F1 score with XGBoost and CatBoost classifiers.
- Wine Quality Classification: Developed and optimized a classification model to predict wine quality based on a Kaggle dataset. Achieved a prediction accuracy of 75.11%.
- Time Series Forecasting:
- Gold Price Prediction: Developed and evaluated models to predict the next day's gold price. Linear Regression had the best performance (MAE: 12.40 USD), with LSTM also performing well. Highlighted the challenges of financial market prediction and the need for additional features for better accuracy.
- Natural Language Processing (NLP):
- Machine Translation (English > German): Experimented with a hand-coded Transformer-based model for translating English text to German. Achieved a BLEU score of 0.0149 on the validation set.
- SMS Spam Detection: Developed a classification system to detect spam messages in SMS text data using Naive Bayes classifiers and text vectorization methods. Achieved a classification accuracy of 98% on the test set.
- Computer Vision projects: Object Detection, Image Segmentation, Image Generation
- Natural Language Understanding (NLU) projects: Chatbots, Sentiment Analysis, Text Classification, Intent Recognition, Large Language Models (LLMs)
- Deep Learning Research: Exploring new architectures, optimization techniques and applications of deep learning.
Feel free to reach out to me via email if you are interested in collaborating or have any questions!