Setting up Focus Group
Read More Project Repo
Documentation and Coding Repo
- Focused Group is created for different Projects.
- Group leader(s) will be assigned for proper managemant and documaentation of the projects in GitHub.
- Group members will work in several sections of the projects and contribute equally. Multiple Jupyter notebooks will be created and managed in the assigned GitHub repo.
- Each group will bring update wikely or biweekly. Each group will be supervised by NPS School of Computing group.
Problem Identification
Read More Project Repo
From Innovative Topics
- Each Focus group will follow one of the subject syllabus (e.g., Computational Neuroscience, Complex System and Network, Computational Genomics, Econophysics, Machine Learning, Deep Learning, Quantum Computing) and figure out the problems that can be translated into coding problem
- Each problem will be organized in a single Jupyter notebook with sufficient documentation (e.g.,title, subtitle, description, mathematical equation, algorithm design, seudocode, coding section and enough visualization)
- Focused group member can participate in multiple problems under the leadership of group leader(s) and supervision of group supervisor. Goup leader(s) will take care of proper management of all problem selected.
Model Development and Analysis
Read More Project Repo
Library and Data management
- Mojority of the problems in these innovative topics have well organized database and python libraries (e.g., Gene Database, Brain2, TensorFlow, Sklearn, networkX, Neo4J )
- Problems in Machine Learning are data driven model development, which required a well defined data set. Problems in Neuroscience are based on Numerical method and simulation while from Complex Network are graph data (e.g., from social media, corruption network).
- Each Focus group will try to implement Numerical Method, simulation, optimization, machine learning technique in their project whenever possible. Introductory concepts in these topics are delivered in Introductory Python training course as well as additional technique will be delivered during the project.
Results and Visualization
Dataframe, Data Plotting 2D, 3D, and Animation
- Majority of the problems in these generates sufficient intermediate data (e.g., array, matrix, vector, functions, 3D surface) for visualization. Visualization of data offers a distinct opportunity to descibe the problem in figure or movie.
- Several introductory (pylab, matplotlib, seaborn, plotly) and advanced datavisualization (d3.js, tree.js, OpenGL, WebGL) tchniques will be explored for various type of problems in physics
- Some problem can be developed as wedget for interactive data plotting and visualization (e.g., changing parameter value using linear scale or value input window)
- Simple time evolution, simulation or roated reference frame can be implemented to create classroom animation (GIF) by collecting number of timeline plots and rotated plots.