Delivery

1. Code
2. Documentation
3. Output

Code

• Either .ipynb or .py
• requirements.txt - List of dependent libraries
• References, if any code is reused

Documentation

• Document (.md or .rst) how to
  • Provide input
  • Run
  • Collect Output.
• Documentation (.md or .rst) of the output
• Maximum of 5-Min video of running the code and generating the output – with other description
  • Use zoom with screen-share and record to cloud to create this video and send the link.

Output

• Create separate folders for each node.
  • Minimum: 1-Node
• In each node-folder
  • Create folders for each metrics and place generated files in the these 4 folders.
    • CPU (At least 1 of the below three)
      • percent-user
      • percent-system
      • percent-idle
    • Memory (At least 1 of the below two)
      • used
      • free
    • Interface (At least 1 of the below two)
      • Packets/Octets
      • Dropped/Errors
    • Load
      • load*
• Each files should have at least 7000 Entries.
• * Only load file will have more than 2 columns.
• Zip the main folder
• Name it with your team name.

Assessment

Categories

1. Metrics Generated
   1. CPU, Memory, Network and Load.
2. Novelty
   1. Neural Network
   2. Discriminator
3. Accuracy
4. Range Validity
   1. Max and Mins
   2. Variations
   3. Trend
5. Implementation
   1. Code Quality
   2. Code Re-Use
6. Individual Metrics
   1. Distribution
   2. Autocorrelation
   3. ARIMA
7. Comparative Metrics
   1. DTW
   2. Wasserstein Distance
   3. RMSE
   4. Maximum Mean Discrepancy
   5. Mutual Information

Weightage Distribution