Motivated Business Intelligence Engineer 1 (Data Scientist) with 1.7 years of progressive experience. I have a strong background in mathematics and expertise in various fields such as machine learning, deep learning, computer vision, natural language processing (NLP) and statistics.
1. Product Reviews Insights
• Designed and implemented a pipeline utilizing AWS Tools to handle a substantial volume, approximately 3TB of raw reviews data. This pipeline efficiently transformed the data into actionable insights through Machine learning, further processed and ingested in a data lake.
• Experimented with various machine learning algorithms such as XGBoost, LightGBM, and Bert (Bidirectional Encoder Representations from Transformers) Model for text classification. Considering cost effectiveness, I implemented LightGBM ML algorithm in AWS Lambda.
• An SQL query was written to extract KPIs and the processed data for visualization. A visualization dashboard was created using AWS Quick sight.
• Created a Scalable solution that could be expanded to other marketplaces.
Tools Used: Python, SQL, Multi-threading, CloudWatch, IAM Role, AWS Lambda, Cloud9, S3, AWS Step Function, ETL, Machine learning, NLP, AWS Quick sight, Redshift.
2. Anomaly detection
• Worked closely with stakeholders to understand the business problem, use cases and requirement.
• Different Statistics technique was applied to explore the relationship between different attributes and identify any correlation.
• Experimented with various Machine learning algorithms like Isolation Forest and One class SVM to identify Anomaly.
• Developed a self-service tool to identify Anomaly in the dataset using AWS Tools.
Tools Used: Python, SQL, S3, AWS Lambda, Amazon Sage maker, Machine learning.
Programming Language
Python, SQL
Tools and Libraries
Scikit-Learn, NumPy, Pandas, Seaborn, Matplotlib, OpenCV, PyTorch, NLTK, SpaCy,multi-threading, Latex, Microsoft Office
AWS Serives
CloudWatch,IAM Role,Cloud watch,AWS Lambda,Step function,Amazon Sage Maker,Cloud9 (Basics),S3 Bucket,AWS Quick sight
GitHub
• Few M.L algorithms and Image processing algorithms (Geometric transforms, Occlusion detection, blurring techniques, Shape from focus (2D to 3D), Othu’s thresholding) from scratch was implemented in Python
Few M.L algorithms and Image processing algorithms (Geometric transforms, Occlusion detection, blurring techniques, Shape from focus (2D to 3D), Othu’s thresholding) from scratch was implemented in Python
1. M. Tech Project (2021-2022)
Generation of Orbital Angular Momentum (OAM) Beams and de-multiplexing OAM using Convolutional neural network (CNN).
• Researched OAM and its demultiplexing methods.
• Python script was written to generate different OAM Modes. Data augmentation was performed on the images to increase the training dataset size.
• Various CNN architectures, such as AlexNet and VGG16, were utilized to enhance accuracy through fine-tuning hyperparameters using Wandb.
2. Introduction to Machine learning Course Project (Jan–May 2021)
Hand Gesture Recognition using Machine learning.
• The main objective is to develop the Accelerometer and Gyroscope based Gesture recognition algorithm using Machine learning.
• First, the acceleration and Gyroscope data of 10 gestures was collected (MPU6050 sensor was used).
• For the following acceleration and Gyroscope data different M.L algorithms was applied, to achieve high accuracy in gesture classification in real time.