Skip to content

GitLab

  • Menu
Projects Groups Snippets
    • Loading...
  • Help
    • Help
    • Support
    • Community forum
    • Submit feedback
    • Contribute to GitLab
  • Sign in / Register
  • ketos ketos
  • Project information
    • Project information
    • Activity
    • Labels
    • Planning hierarchy
    • Members
  • Repository
    • Repository
    • Files
    • Commits
    • Branches
    • Tags
    • Contributors
    • Graph
    • Compare
  • Issues 27
    • Issues 27
    • List
    • Boards
    • Service Desk
    • Milestones
  • Merge requests 0
    • Merge requests 0
  • CI/CD
    • CI/CD
    • Pipelines
    • Jobs
    • Schedules
  • Deployments
    • Deployments
    • Environments
    • Releases
  • Monitor
    • Monitor
    • Incidents
  • Analytics
    • Analytics
    • Value stream
    • CI/CD
    • Repository
  • Wiki
    • Wiki
  • Snippets
    • Snippets
  • Activity
  • Graph
  • Create a new issue
  • Jobs
  • Commits
  • Issue Boards
Collapse sidebar
  • public_projects
  • ketosketos
  • Merge requests
  • !55

Merged
Created Oct 22, 2018 by Oliver Kirsebom@kirsebomOwner2 of 2 tasks completed2/2 tasks

Refactor neural network

  • Overview 1
  • Commits 19
  • Pipelines 17
  • Changes 9

This merge request provides the following improvements/enhancements:

  1. The class CNNWhale is split into a parent class called DataHandler, and a child class called CNNWhale, where DataHandler provides generic implementations of methods related to data handling (set, add, get training/validation/test data).

  2. Some method names have been simplified (e.g. create_net_structure -> create, save_model -> save, etc.)

  3. A new neural network class called 'EDTCN' (Encoder-Decoder Temporal Network) has been implemeted using keras methods from the Tensorflow library.

  4. A new class called TrainingDataProvider has also been implemented. This class facilitates an active learning process, in which extra emphasis is placed on the training examples that yield low confidence predictions or wrong predictions.

Things still to be done:

  1. Implement more unit tests of TrainingDataProvider
  2. Finish implementation of EDTCN (including unit tests)
Edited Nov 28, 2018 by Oliver Kirsebom
Assignee
Assign to
Reviewer
Request review from
Time tracking
Source branch: refactor_neural_net