Package to apply MTL on a few dataset
Alternatives To Multitasklearning
Project NameStarsDownloadsRepos Using ThisPackages Using ThisMost Recent CommitTotal ReleasesLatest ReleaseOpen IssuesLicenseLanguage
Sign Language Digits Dataset179
2 years agoapache-2.0
Turkey Ankara Ayrancı Anadolu High School's Sign Language Digits Dataset
Exams Qa23
a year agocc-by-sa-4.0Python
A Multi-subject High School Examinations Dataset for Cross-lingual and Multilingual Question Answering
3 days ago35otherPython
Canadian COVID-19 Data Archive
a year ago1agpl-3.0C#
🏫 🇧🇩 Fuzzy-AHP-based recommendation system for secondary schools in Bangladesh
6 years agoPython
Package to apply MTL on a few dataset
Naplan Results Reporting10
10 months agoPerl
NAPLAN Results and Reporting draft dataset
3 years ago2mitCSS
Project for helping professors with grading.
7 years agoJupyter Notebook
Datasets and Matplotlib/pandas examples
5 years ago
source code and datasets for PCM2018
4 years ago1otherR
An R Package to provide an easy API to the disparate UK Education Datasets. Includes Get Schools Information (Edubase) at present
Alternatives To Multitasklearning
Select To Compare

Alternative Project Comparisons

#Multi-Task Learning Authors : Charles Corbière, Hamza Cherkaoui


This package implement differents multi-task learning models:

  • Multilearning SVM (svm): an SVM is learning for each task
  • Alternating Structure Optimization (aso): a modele assuming every task shared a low dimensional structure
  • Convex Alternating Structure Optimization (caso): convex relaxation of ASO
  • Clustered Multi-task Learning (cmtl): a modele assuming tasks are groupd within clusters.

Dataset included:

  • a clustered toy dataset (toy)
  • School data (school)
  • Sarcos data (sarcos)

How to use it

  • To compute score for a given algorithm on a given dataset, for a test size proportion and a number of splits
python school cmtl 5 0.30

Here, we run 5 times CMTL on school dataset with a 30% test size proportion.

  • To plot all algorithms scores for a given dataset and a number of splits, iterating on the test size proportion
python school 5

Here, we run 5 times for each algorithm on school dataset. Note that on current implementation, the test size range is [0.30, 0.40, 0.50, 0.60]

Popular Dataset Projects
Popular School Projects
Popular Data Processing Categories

Get A Weekly Email With Trending Projects For These Categories
No Spam. Unsubscribe easily at any time.