Sayemuzzaman Siam

GitHub | LinkedIn | sayemuzzaman505@gmail.com | Medium | Kaggle | MonkeyType
Phone: +8801839592354

Sayemuzzaman Siam

About Me

Motivated Computer Science and Engineering graduate with hands-on experience in deep learning, neural networks, and data-driven projects. Proficient in implementing innovative solutions in computer vision, natural language processing, and predictive modeling. Passionate about Deep Learning and enjoy researching to contribute to the field of Computer Vision.

Education

East West University

Bachelor of Science in Computer Science and Engineering

CGPA: 3.04/4.00

Relevant Coursework: Computer Networks, Data Structures and Algorithms, Artificial Intelligence, Web Programming, Machine Learning, Digital Image Processing, Software Engineering, Statistics for Data Science.

Technical Skills

Projects

Image Caption Generators

Designed Image Captioning models using CNNs (VGG16, ResNet50, EfficientNet-B3) and LSTM on Flickr8k and Flickr30k. Evaluated the models using BLEU Scores.

Technologies: TensorFlow, Keras, NumPy, Pandas, CNNs

GitHub: Image Caption Generators

Image Classifications

Dog vs Cat Image Classification: Created it using both Custom CNN Model and Pre-Trained model on dogs vs cats dataset. The Custom CNN was able to achieve 80% accuracy while the Pre-Trained model MobileNetV2 achieved 98% accuracy.

Technologies: TensorFlow, Keras, CNN, Transfer Learning

GitHub: Image Classifications

Pneumonia Detection in Chest X-Rays and Explainable AI

Designed a deep learning model to classify chest X-ray images for pneumonia detection on chest x-ray dataset. Incorporated Explainable AI techniques (Grad-CAM, Lime, Integrated Gradients) to interpret the model's decisions and transparency.

Technologies: Python, TensorFlow, Keras, XAI, NumPy, EfficientNetB0

GitHub: Computer Vision with XAI

Sentiment Analysis from Text

Developed a natural language processing (NLP) model to classify emotions on ten classes in English text using LSTM.

Technologies: Python, TensorFlow, LSTM

GitHub: Sentiment Analysis

Big Mart Sales Prediction

Developed a regression model to predict sales based on historical data.

Technologies: Python, Pandas, NumPy, Scikit-learn, Matplotlib, Seaborn

GitHub: Sale Prediction

Flickr8k Data Analysis

Conducted data analysis on Flickr8k dataset to get hidden insights.

Technologies: Python, Pandas, Matplotlib, Seaborn

GitHub: Flickr8k Analysis

Recommendation System

Build Book and Movie recommendation system using popularity based filtering, collaborative filtering and content-based approaches. Used Book-Crossing: User review ratings and MovieLens 20M Datasets.

Technologies: Python, Pandas, Numpy, Cosine Similarity

GitHub: Recommender-Systems

Research Publication

Explainable Artificial Intelligence for Deep Learning Based Detection of Pneumonia in Chest X-ray Images

Presented at: 3rd International Conference on Innovations in Data Analytics (ICIDA 2024)

Date: 18th-19th December 2024

Organized by: Eminent College of Management & Technology (ECMT)