Farjad Ahmed Siddiqui

I am a data scientist with a strong analytical mindset and a passion for solving complex problems using data-driven insights. With a background in data science and machine learning, I have experience in data preprocessing, exploratory data analysis (EDA), clustering, and predictive modeling. I have worked on various projects, including:

Clustering Analysis: Applied K-Means, DBSCAN, OPTICS, and Gaussian Mixture Models (GMM) on real-world datasets, optimizing clustering performance through data transformations and parameter tuning.

Neural Networks: Built and fine-tuned feed-forward neural networks from scratch, implementing forward and backward propagation for binary classification tasks.

LLM & NLP Projects: Compared different LLM models, fine-tuned Tiny-BERT for sentiment analysis on IMDb reviews, and built a Retrieval-Augmented Generation (RAG) chatbot for migration-related queries.

BI & Time Series Forecasting: Developed BI solutions and applied ARIMA, SARIMA, and Exponential Smoothing models for time series forecasting using real-world temperature datasets.

Computer Vision: Implemented vehicle speed estimation using YOLOv8 and the supervision library.

I am proficient in Python, SQL, and machine learning frameworks like TensorFlow, PyTorch, and Hugging Face. My experience extends to working with data visualization, feature engineering, and cloud-based solutions like Kaggle’s GPU environment.

I am actively seeking opportunities where I can contribute my skills to impactful projects, collaborate with cross-functional teams, and continue learning in the fields of AI, data science, and automation.