166 Hits in 4.8 sec

Interpretable MTL from Heterogeneous Domains using Boosted Tree [article]

Ya-Lin Zhang, Longfei Li
2020 arXiv   pre-print
What's more, interpretability can be naturally obtained from the tree based method, satisfying the industrial needs.  ...  However, in industrial scenarios, interpretability is always demanded, and the data of different tasks may be in heterogeneous domains, making the existing methods unsuitable or unsatisfactory.  ...  CONCLUSIONS This paper proposes a boosted tree based MTL method for heterogeneous tasks. A two-stage method is developed with many strategies (i.e., regularization and early stopping) used.  ... 
arXiv:2003.07077v1 fatcat:txbdze63rjbi5fbqqgo4kfhz7e

Task-wise Split Gradient Boosting Trees for Multi-center Diabetes Prediction [article]

Mingcheng Chen, Zhenghui Wang, Zhiyun Zhao, Weinan Zhang, Xiawei Guo, Jian Shen, Yanru Qu, Jieli Lu, Min Xu, Yu Xu, Tiange Wang, Mian Li (+5 others)
2021 arXiv   pre-print
To tackle the above challenges, we employ gradient boosting decision trees (GBDT) to handle data heterogeneity and introduce multi-task learning (MTL) to solve data insufficiency.  ...  To this end, Task-wise Split Gradient Boosting Trees (TSGB) is proposed for the multi-center diabetes prediction task.  ...  An MTL algorithm with boosted decision tree is proposed in [10] , but it is designed for heterogeneous MTL i.e., the tasks share the same data but with different labels set for each task.  ... 
arXiv:2108.07107v1 fatcat:ohwivx56lngihgn3ozxfgn3nxe

Phenotyping with Prior Knowledge using Patient Similarity

Asif Rahman, Yale Chang, Bryan Conroy, Minnan Xu-Wilson
2020 Machine Learning in Health Care  
We present a method to learn from prior knowledge using a mixture-of-experts model where gating probabilities are tuned by an adjacency matrix created using side information available during training,  ...  , and extracted from knowledge graphs.  ...  Embedding Decision Paths from Boosted Trees The expert model we used combines gradient boosted decision trees (GBDT) with neural networks.  ... 
dblp:conf/mlhc/RahmanCCX20 fatcat:5emhm5q3hzckhcsysaqsxy6kve

Maintenance intervention predictions using entity-embedding neural networks

Zaharah Allah Bukhsh, Irina Stipanovic, Aaqib Saeed, Andre G. Doree
2020 Automation in Construction  
The subjective approach is likely to lack the appropriate use of inspection data and does not promise cost-effective maintenance plans.  ...  In this paper, we show that the historical and operational data, readily available at the agencies, is of vital importance and can be used effectively for the recommendations of maintenance advises for  ...  Acknowledgement This study has been performed under funding from the European Union's Horizon 2020 -Research and Innovation Framework Programme with grant agreement No 636285 DESTination Rail.  ... 
doi:10.1016/j.autcon.2020.103202 fatcat:jhxa5ir2a5dw5ncerjbxio2om4

A Survey on Multi-Task Learning [article]

Yu Zhang, Qiang Yang
2018 arXiv   pre-print
Many real-world applications use MTL to boost their performance and we review representative works. Finally, we present theoretical analyses and discuss several future directions for MTL.  ...  Multi-Task Learning (MTL) is a learning paradigm in machine learning and its aim is to leverage useful information contained in multiple related tasks to help improve the generalization performance of  ...  Without special explanation, the default MTL setting is the homogeneous-feature MTL. Here we need to distinguish the heterogeneous-feature MTL from the heterogeneous MTL.  ... 
arXiv:1707.08114v2 fatcat:6lrpe4nk45djbjyfjco7t4yfme

The HARMONIA Project: Hardware Monitoring for Automotive Systems-of-Systems [chapter]

Thang Nguyen, Ezio Bartocci, Dejan Ničković, Radu Grosu, Stefan Jaksic, Konstantin Selyunin
2016 Lecture Notes in Computer Science  
Observers embedded on FPGA hardware will be generated from assertions, and used for monitoring automotive designs emulated on hardware.  ...  In such scenario, any effort to reduce the design and verification costs and to improve the time-to-market and the product quality will play an important role to boost up the competitiveness of the automotive  ...  This research is supported by the project HARMONIA (845631), funded by a national Austrian grant from FFG (Österreichische Forschungsförderungsgesellschaft) under the program IKT der Zukunft.  ... 
doi:10.1007/978-3-319-47169-3_28 fatcat:fwwhfmjp3nhljbl5rn2udf2nuy


Sirat Samyoun, Md Mofijul Islam, Tariq Iqbal, John Stankovic
2022 Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies  
We present M3Sense, a multi-task, multimodal representation learning framework that effectively learns the affect-agnostic physiological representations from limited labeled data and uses a novel domain  ...  alignment technique to utilize the unlabeled data from the other affective tasks to accurately detect these mental health conditions using wrist sensors only.  ...  data from heterogeneous domains through the domain alignment module.  ... 
doi:10.1145/3534600 fatcat:oyt4lh2nqzglbivw5mersvd3ea

Multi-Task and Multi-Corpora Training Strategies to Enhance Argumentative Sentence Linking Performance [article]

Jan Wira Gotama Putra and Simone Teufel and Takenobu Tokunaga
2021 arXiv   pre-print
In this paper, we improve a state-of-the-art linking model by using multi-task and multi-corpora training strategies.  ...  We would like to thank anonymous reviewers for their useful and detailed feedback.  ...  The F1-macro for individual link predictions was boosted to .374 from .323.  ... 
arXiv:2109.13067v1 fatcat:owqh6mxclzfmpftu4hm3aqxmkq

Framework for Deep Learning-Based Language Models using Multi-task Learning in Natural Language Understanding: A Systematic Literature Review and Future Directions

Rahul Manohar Samant, Mrinal Bachute, Shilpa Gite, Ketan Kotecha
2022 IEEE Access  
Even though MTL (Multi-task Learning) was introduced before Deep Learning, it has gained significant attention in the past years.  ...  INDEX TERMS Deep learning, Knowledge representation, Multi-task NLU, Unsupervised learning TABLE I APPLICATION DOMAINS FOR NLU Domain Applications Machine translation IBM Watson Task-based dialogue-based  ...  Text data belongs to heterogeneous sources like social media, electronic communication, or interrogative data like QA from the client interaction.  ... 
doi:10.1109/access.2022.3149798 fatcat:k3kdt4eryzdfpk5k6w62jtlskm

A Multitask Learning Model for Traffic Flow and Speed Forecasting (April 2020)

Kunpeng Zhang, Lan Wu, Zhaoju Zhu, Jiang Deng
2020 IEEE Access  
Intelligent Transportation Systems (ITS) research and applications benefit from accurate shortterm traffic state forecasting.  ...  The results suggest the proposed MTL-GRU model with residual mappings is promising to forecast short-term traffic state.  ...  A set of past values of variables (i.e., traffic flow or traffic speed) are selected and fed into the corresponding v-SVM model. 4) XGBoost: XGBoost is a scalable machine learning system for tree boosting  ... 
doi:10.1109/access.2020.2990958 fatcat:urk2fun3c5d3lc7gnzgkl3yily

Location-Centered House Price Prediction: A Multi-Task Learning Approach [article]

Guangliang Gao, Zhifeng Bao, Jie Cao, A. K. Qin, Timos Sellis, Fellow, IEEE, Zhiang Wu
2019 arXiv   pre-print
We propose a location-centered prediction framework that differs from existing work in terms of data profiling and prediction model.  ...  We address this problem by conducting a careful study of exploiting the Multi-Task Learning (MTL) model.  ...  Decision tree is often used as the base learner of the ensemble. In this study, the base learner is actually a regression tree.  ... 
arXiv:1901.01774v1 fatcat:xfe5zplcfrbl7i6xxms4jtwhbe

Transfer Learning-Based Outdoor Position Recovery with Telco Data [article]

Yige Zhang, Aaron Yi Ding, Jorg Ott, Mingxuan Yuan, Jia Zeng, Kun Zhang, Weixiong Rao
2019 arXiv   pre-print
(those fine-grained small subareas) to a target one which originally suffers from poor localization accuracy.  ...  Specifically, TLoc introduces three dedicated components: 1) a new coordinate space to divide an area of interest into smaller domains, 2) a similarity measurement to select best source domains, and 3)  ...  This greatly boosts the performance in the final model, at the expense of a small increase in the bias and loss of interpretability.  ... 
arXiv:1912.04521v1 fatcat:bdi4nzcr3jgizgvlhlu2tv4kbe

Selecting Optimal Trace Clustering Pipelines with AutoML [article]

Sylvio Barbon Jr, Paolo Ceravolo, Ernesto Damiani, Gabriel Marques Tavares
2021 arXiv   pre-print
Trace clustering has been extensively used to preprocess event logs.  ...  Our experiments were conducted using a thousand event logs, four encoding techniques, and three clustering methods.  ...  Event logs and featurization MtL benefits from using a large set of instances in the meta-database.  ... 
arXiv:2109.00635v1 fatcat:jvvqi7wbqjggnbebjjrtzkmory

DrugOrchestra: Jointly predicting drug response, targets, and side effects via deep multi-task learning [article]

Yuepeng Jiang, Stefano Rensi, sheng wang, Russ B. Altman
2020 bioRxiv   pre-print
We constructed a heterogeneous drug discovery dataset of over 21k drugs by integrating 8 datasets across three tasks.  ...  Instead of directly fine-tuning on an individual task, DrugOrchestra uses deep multi-task learning to obtain a phenotype-based drug representation by simultaneously fine-tuning on drug response, target  ...  Qiao et al. applied an MTL model to a heterogeneous dataset integrated from 12 different individual cancer drug response datasets [27] .  ... 
doi:10.1101/2020.11.17.385757 fatcat:un3gsz4l2bfpfbqgfgglkwqbxe

Multi-study inference of regulatory networks for more accurate models of gene regulation [article]

Dayanne M. Castro, Nick de Veaux, Emily R. Miraldi, Richard Bonneau
2018 bioRxiv   pre-print
In addition, adaptive penalties may be used to favor models that include interactions derived from multiple sources of prior knowledge including orthogonal genomics experiments.  ...  We evaluate generalization and network recovery using examples from Bacillus subtilis and Saccharomyces cerevisiae, and show that sharing information across models improves network reconstruction.  ...  a better approach to integrate datasets from 211 different domains.  ... 
doi:10.1101/279224 fatcat:jjnbpuwczbdm5iqleaxxacprby
« Previous Showing results 1 — 15 out of 166 results