Questions tagged [multitask-learning]

Multi-task learning (MTL) is an approach to machine learning that learns a problem together with other related problems at the same time, using a shared representation.

Filter by
Sorted by
Tagged with
1
vote
1answer
7 views

Parallel multi-task learning vs. continual learning

Assuming we want to learn k tasks jointly, and the data for all tasks are available. We may either train a model with parallel multi-task learning (eg. each batch is a mixture of samples from the k ...
0
votes
1answer
50 views

Multi-task learning: weight selection for combining loss functions

I am training a system that combines two sub-systems: one for classification and another for reconstruction. Can anyone suggestion what are the common practice for weight selection for combining two ...
1
vote
0answers
53 views

Covariance Matrix of HIERARCHICAL MULTITASK GAUSSIAN PROCESS

I'm currently trying to develop a Gaussian Process to predict different levels of different individuals over time. So it is a Time Regression Problem in which we have multiple tasks, but also ...
0
votes
0answers
20 views

Demand forecasting for sequence data and multi-task learning

I have a demand forecasting problem that I'd like to solve with a deep learning using multi-task learning and I'd like advice in some areas. Problem definition: I have a set of $N$ customers that ...
2
votes
0answers
162 views

How can I maximise binary cross entropy loss?

I have a multi-task learning model with two binary classification tasks. One part of the model creates a shared feature representation that is fed into two subnets in parallel. The loss function for ...
2
votes
1answer
82 views

Classification followed by regression

I have the following problem: I have a dataset for which my observations have a bunch of features and a continuous response (regression problem). However, some of my observations (about a fourth of ...
4
votes
1answer
423 views

Multi-task XGBoost

Is there a way to adapt the XGBoost algorithm to the multi-task case? Say there are related output variables and for some samples, some of those outcomes are missing. Is there a way to train XGBoost ...
1
vote
1answer
1k views

How to define multiple losses in machine learning?

I'm using TensorFlow for training CNN for classification. In machine learning, there are several different definitions for loss function. In general, we may select one specific loss (e.g., binary ...
2
votes
0answers
120 views

How to optimize the hyperparameters of a Deep Gaussian Process?

I am trying to understand an article from NIPS 2017 where Alaa and van der Schaar create a Deep Multitask Gaussian Processes (DMGP) with competing risks. I don't grab what are they trying to optimize. ...
1
vote
1answer
31 views

Robust machine learning for slightly different class proportions in multiple data sets

Say we have n similar data sets, with the same variables, and outcome labels x and y. In these data sets, domains slightly differ as suggested by the proportion of the minority class x (ranging from 1%...
1
vote
0answers
757 views

How to add task wise early stopping in keras?

I created a multi task network with keras. Because different tasks have different level of difficulties, for some tasks overfitting occurs earlier than others so they should be stopped. I know keras ...
0
votes
2answers
95 views

Multitask learning Gaussian Processes

A small question In the book of Rasmussen Page 115 last paragraph. When we have multiple databases you setup a gaussian for each database and the optimisation is said can be done by adding the ...
5
votes
2answers
4k views

Difference between multitask learning and transfer learning

I am reading Caruana (1997) Multitask learning (pdf). In the definition of multi task learning, the author states that; Usually, we do not care how well extra tasks are learned; their sole purpose ...
0
votes
2answers
95 views

How to apply the SVD based approach to sample from a matrix-variate normal with given correlation matrices and standard deviations?

A sample from a multivariate normal distribution $X$ can be constructed to have a covariance $C$ even for positive semi-definite covariances according to this technique involving an SVD. Furthermore, $...
3
votes
2answers
573 views

Sampling from matrix-variate normal distribution with singular covariances? [duplicate]

The matrix-variate normal distribution can be sampled indirectly by utilizing the Cholesky decomposition of two positive definite covariance matrices. However, if one or both of the covariance ...
3
votes
1answer
151 views

Training N classifiers for N labels vs one classifier with N labels

I have a classification problem which is multi-label with N labels. I would like to know which method would be the better choice? Training N classifiers (1 for each label) or a single classifier which ...
1
vote
0answers
264 views

What are 'tasks' in multi-task/transfer learning, linking to co-regionalisation

I am currently working on implementing a multi-task model for Gaussian process prediction, but am having some trouble interpreting the materials online (conference talks, papers, presentations/slides ...
0
votes
0answers
76 views

Modeling worker performance parameters for optimum allocation of tasks to workers

Lets say we have an English to French translation task in a company and there are 100s of workers who are proficient in doing this task but each worker has its own unique attributes which allow them ...
6
votes
4answers
2k views

What is the difference between Multitask and Multiclass learning

Consider a image labeling problem, where I need to assign one or more labels to an image. The possible labels are human, moving ,...
0
votes
1answer
46 views

In the context of ANNs, is there multi-task learning iff the network has more than 1 output?

This is a terminology question: in the context of artificial neural networks, does multi-task learning occur iff the network has more than 1 output?