-
Notifications
You must be signed in to change notification settings - Fork 0
Worksheet: Jan
Jan Svačina edited this page Oct 25, 2019
·
7 revisions
| Week | Hours | Issue | Description | Evidence |
|---|---|---|---|---|
| Week 6. | 6 | (SCA11), (SCA20) | Developing the idea and preparing proposed method | |
| Week 7. | 7 | (SCA12) | Preparing presentation and documentation, 1. phase, adding comments | presentation |
| Week 8. | 5 | (SCA16) Entropy Detection | Exteded project to be able to use different algorithms, for instance entropy for our evaluation method | e04fb2e |
| 8 | Preparing presentation and documentation, 2. phase, adding comments | presentation | ||
| SUM | 26 |
| Code | Name | Description | Asigned To | Status |
|---|---|---|---|---|
| SCA-20 | Filters / Convolutions for nested classifications | We need to have some sort of convolution or filter on the input vectors to be able to recognize not only if a line of code is 80% loop and 20% conditional, for example, but also that if it were a loop as it suggests that it is say 95% for loop, 3% while, 2% do while, etc. Essentially: investigate methods of convolution and creating deeper classifications. | Jan | Backlog |
| SCA-12 | Discover a worthy DL Model | We need to have a deep learning model that supports our system. Currently we are using 3 layered sequential network with 64 nodes each and are only comparing tokens | Jan | In Development |
| SCA-16 | Entropy Detection | Given code that has been classified line by line, detect the entropy with experimental weights | Jan | In Development |
| SCA-11 | Creating a proposed classification model | We need to have distinct categories for classifying lines of code | Jan | Review |
- creating proposed method and creating presentation with the idea
- 1 git commits
- many hours of learning how to use tensorflow with Python
- many hours brainstorming with Jonathan