• 🔭 I’m currently working on Computer Vision and Predictive Modeling techniques

• 👨‍💻 All of my projects are available at https://github.com/ShivSubedi

• 📫 How to reach me [email protected]

• Developed a CNN-based solution to characterize the spatial resolution of a novel semi-monolithic PET detector geometry, leveraging 1600 training and 1521 testing points (1mm pitch). This work demonstrates my ability to apply deep learning to complex medical imaging challenges.
• Engineered a novel method for correcting position bias in CNN predictions, significantly improving the accuracy and reliability of the model. This is critical for precise image reconstruction in PET scans and showcases my skills in model optimization and performance analysis.
• First-author publication accepted in Radiation Measurements (pre-print available on arXiv: link). This publication highlights my experience building and optimizing CNN models, including developing simple yet robust architectures and training strategies.

You can find rest of my research publications and citations on my Google Scholar profile.

I possess a strong foundation in data science and machine learning, enabling me to develop and deploy effective solutions. My key skills include:

• Data Analysis & Manipulation: Proficient in using Python libraries like NumPy and Pandas to clean, process, and analyze large datasets.

• Machine Learning Model Development: Experienced in building and training machine learning models using frameworks such as TensorFlow and PyTorch. Familiar with various algorithms, including:

  • Supervised Learning: Regression, Classification (e.g., linear regression, logistic regression)
  • Unsupervised Learning: Clustering (e.g., K-means clustering)
  • Deep Learning: Proficient in developing and deploying deep learning models, including:
    • CNNs (Convolutional Neural Networks): Developing CNN architectures for image classification, and object detection.
    • UNet: Experience with UNet architectures for image segmentation, particularly in applications like medical imaging. Applied UNet to liver and lung tumor segmentation in CT images.
    • ResNet: Experience with ResNet architectures for image classification tasks in medical imaging. Applied ResNet to pneumonia detection in X-ray images.

• Computer Vision: Proficient in using OpenCV for image processing, object detection (using pre-trained models), and other computer vision tasks. Experience integrating OpenCV with deep learning models for object detection applications.

• Data Visualization: Able to create insightful visualizations using tools like Matplotlib and Seaborn to communicate findings effectively. Experience with creating various types of charts and graphs to represent data and model performance.

• Software Development & Collaboration: Skilled in using Git for version control and collaborating with teams using platforms like Jira and Confluence.

• Programming Languages:

• Machine Learning Frameworks:

• Cloud-Based IDEs:

• Data Science Libraries/Tools:
  

• AI:

• Data Visualization:

• Database Management:

• Project Management & Collaboration:

• GPU Computing: