2 code implementations • 7 Apr 2024 • Muhammad Tanzil Furqon, Mahardhika Pratama, Lin Liu, Habibullah, Kutluyil Dogancay
MDAN encompasses a three-staged mechanism where the mix-up strategy is not only performed to regularize the source and target domains but also applied to establish an intermediate mix-up domain where the source and target domains are aligned.
1 code implementation • 19 Feb 2024 • Marcus de Carvalho, Mahardhika Pratama, Jie Zhang, Chua Haoyan, Edward Yapp
In this work, we introduce a novel approach called Cross-Domain Continual Learning (CDCL) that addresses the limitations of being limited to single supervised domains.
1 code implementation • 25 Jan 2024 • Muhammad Anwar Ma'sum, Md Rasel Sarkar, Mahardhika Pratama, Savitha Ramasamy, Sreenatha Anavatti, Lin Liu, Habibullah, Ryszard Kowalczyk
A slow learner tailors suitable representations to fast learners.
1 code implementation • 25 Jan 2024 • Naeem Paeedeh, Mahardhika Pratama, Muhammad Anwar Ma'sum, Wolfgang Mayer, Zehong Cao, Ryszard Kowlczyk
Most few-shot learning works rely on the same domain assumption between the base and the target tasks, hindering their practical applications.
1 code implementation • 26 Jun 2023 • Muhammad Anwar Ma'sum, Mahardhika Pratama, Edwin Lughofer, Lin Liu, Habibullah, Ryszard Kowalczyk
This paper proposes a few-shot continual learning approach, termed FLat-tO-WidE AppRoach (FLOWER), where a flat-to-wide learning process finding the flat-wide minima is proposed to address the catastrophic forgetting problem.
no code implementations • 21 Apr 2023 • Md Rasel Sarkar, Sreenatha G. Anavatti, Tanmoy Dam, Mahardhika Pratama, Berlian Al Kindhi
The proposed model is evaluated for single-step and multi-step WPF, and compared with gated recurrent unit (GRU) and long short-term memory (LSTM) models on a wind power dataset.
1 code implementation • 21 Mar 2023 • Muhammad Anwar Ma'sum, Mahardhika Pratama, Edwin Lughofer, Weiping Ding, Wisnu Jatmiko
This paper proposes an assessor-guided learning strategy for continual learning where an assessor guides the learning process of a base learner by controlling the direction and pace of the learning process thus allowing an efficient learning of new environments while protecting against the catastrophic interference problem.
1 code implementation • 5 Sep 2022 • Marcus de Carvalho, Mahardhika Pratama, Jie Zhang, Yajuan San
Catastrophic forgetting has been a significant problem hindering the deployment of deep learning algorithms in the continual learning setting.
1 code implementation • 4 Sep 2022 • Tanmoy Dam, Mahardhika Pratama, Md Meftahul Ferdaus, Sreenatha Anavatti, Hussein Abbas
Adversarial continual learning is effective for continual learning problems because of the presence of feature alignment process generating task-invariant features having low susceptibility to the catastrophic forgetting problem.
1 code implementation • 4 Sep 2022 • Weiwei Weng, Mahardhika Pratama, Choiru Za'in, Marcus de Carvalho, Rakaraddi Appan, Andri Ashfahani, Edward Yapp Kien Yee
This paper aims to attack the problem of extreme label shortage in the cross domain multistream classification problems where only very few labelled samples of the source stream are provided before process runs.
no code implementations • 4 Sep 2022 • Tanmoy Dam, Md Meftahul Ferdaus, Mahardhika Pratama, Sreenatha G. Anavatti, Senthilnath Jayavelu, Hussein A. Abbass
Many real-world classification problems have imbalanced frequency of class labels; a well-known issue known as the "class imbalance" problem.
no code implementations • 4 Sep 2022 • Appan Rakaraddi, Siew Kei Lam, Mahardhika Pratama, Marcus de Carvalho
Continual learning on graphs is largely unexplored and existing graph continual learning approaches are limited to the task-incremental learning scenarios.
1 code implementation • 8 Nov 2021 • Appan Rakaraddi, Mahardhika Pratama
So, it is of the essence to develop a method that is scalable in low computational time.
1 code implementation • 4 Oct 2021 • Marcus de Carvalho, Mahardhika Pratama, Jie Zhang, Edward Yapp
We consider the problem of online unsupervised cross-domain adaptation, where two independent but related data streams with different feature spaces -- a fully labeled source stream and an unlabeled target stream -- are learned together.
no code implementations • 20 Sep 2021 • Andri Ashfahani, Mahardhika Pratama
A deep clustering network is desired for data streams because of its aptitude in extracting natural features thus bypassing the laborious feature engineering step.
1 code implementation • 5 Sep 2021 • Renchunzi Xie, Mahardhika Pratama
Knowledge transfer across several streaming processes remain challenging problem not only because of different distributions of each stream but also because of rapidly changing and never-ending environments of data streams.
no code implementations • 28 Jun 2021 • Mahardhika Pratama, Andri Ashfahani, Edwin Lughofer
Unsupervised continual learning remains a relatively uncharted territory in the existing literature because the vast majority of existing works call for unlimited access of ground truth incurring expensive labelling cost.
no code implementations • 26 Jun 2021 • Andri Ashfahani, Mahardhika Pratama, Edwin Lughofer, Edward Yapp Kien Yee
The common practice of quality monitoring in industry relies on manual inspection well-known to be slow, error-prone and operator-dependent.
no code implementations • 26 Jun 2021 • Mao Fubing, Weng Weiwei, Mahardhika Pratama, Edward Yapp Kien Yee
Learning from streaming tasks leads a model to catastrophically erase unique experiences it absorbs from previous episodes.
no code implementations • 26 Jun 2021 • Mahardhika Pratama, Choiru Za'in, Edwin Lughofer, Eric Pardede, Dwi A. P. Rahayu
The large-scale data stream problem refers to high-speed information flow which cannot be processed in scalable manner under a traditional computing platform.
no code implementations • 3 Nov 2019 • Mahardhika Pratama, Andri Ashfahani, Mohamad Abdul Hady
The feasibility of existing data stream algorithms is often hindered by the weakly supervised condition of data streams.
2 code implementations • 8 Oct 2019 • Mahardhika Pratama, Marcus de Carvalho, Renchunzi Xie, Edwin Lughofer, Jie Lu
It automatically evolves its network structure from scratch with/without the presence of ground truth to overcome independent concept drifts in the source and target domain.
no code implementations • 8 Oct 2019 • Andri Ashfahani, Mahardhika Pratama, Edwin Lughofer, Yew Soon Ong
The Denoising Autoencoder (DAE) enhances the flexibility of the data stream method in exploiting unlabeled samples.
no code implementations • 8 Oct 2019 • Mahardhika Pratama, Choiru Za'in, Andri Ashfahani, Yew Soon Ong, Weiping Ding
The advantage of NADINE, namely elastic structure and online learning trait, is numerically validated using nine data stream classification and regression problems where it demonstrates performance improvement over prominent algorithms in all problems.
no code implementations • 19 Jul 2019 • Mohamad Abdul Hady, Basaran Bahadir Kocer, Harikumar Kandath, Mahardhika Pratama
At the same time, UAVs may need to operate under external disturbances to follow time-based trajectories.
1 code implementation • 17 Oct 2018 • Andri Ashfahani, Mahardhika Pratama
The feasibility of deep neural networks (DNNs) to address data stream problems still requires intensive study because of the static and offline nature of conventional deep learning approaches.
no code implementations • 24 Sep 2018 • Mahardhika Pratama, Andri Ashfahani, Yew Soon Ong, Savitha Ramasamy, Edwin Lughofer
The generative learning phase of Autoencoder (AE) and its successor Denosing Autoencoder (DAE) enhances the flexibility of data stream method in exploiting unlabelled samples.
no code implementations • 26 Aug 2018 • Mahardhika Pratama, Witold Pedrycz, Geoffrey I. Webb
DEVFNN is developed under the stacked generalization principle via the feature augmentation concept where a recently developed algorithm, namely gClass, drives the hidden layer.
no code implementations • 7 Aug 2018 • Mahardhika Pratama, Dianhui Wang
The concept of SCN offers a fast framework with universal approximation guarantee for lifelong learning of non-stationary data streams.
no code implementations • 18 Jul 2018 • Mahardhika Pratama, Choiru Za'in, Eric Pardede
Extensive experiments on this framework are validated by measuring the accuracy and running time of four combinations of Scalable PANFIS and other Spark-based built in algorithms.
no code implementations • 20 May 2018 • Andri Ashfahani, Mahardhika Pratama, Edwin Lughofer, Qing Cai, Huang Sheng
{Radio Frequency Identification technology has gained popularity for cheap and easy deployment.
no code implementations • 11 May 2018 • Md Meftahul Ferdaus, Mahardhika Pratama, Sreenatha G. Anavatti, Matthew A. Garratt
Data stream has been the underlying challenge in the age of big data because it calls for real-time data processing with the absence of a retraining process and/or an iterative learning approach.
no code implementations • 5 Feb 2018 • Wahyu Caesarendra, Mahardhika Pratama, Tegoeh Tjahjowidodo, Kiet Tieud, Buyung Kosasih
From the results, it is suggested that PANFIS offers outstanding performance compared to those of other methods.
no code implementations • 6 Nov 2017 • Mahardhika Pratama, Eric Dimla, Edwin Lughofer, Witold Pedrycz, Tegoeh Tjahjowidowo
The paper presents advancement of a newly developed ensemble learning algorithm, pENsemble+, where online active learning scenario is incorporated to reduce operator labelling effort.
no code implementations • 29 May 2017 • Murtadha Talib AL-Sharuee, Fei Liu, Mahardhika Pratama
The base classifier of the ensemble method is a modified k-means algorithm.
no code implementations • 18 May 2017 • Mahardhika Pratama, Witold Pedrycz, Edwin Lughofer
pENsemble adopts a dynamic ensemble structure to output a final classification decision where it features a novel drift detection scenario to grow the ensemble structure.
no code implementations • 6 May 2017 • Mahardhika Pratama
We note at least three uncharted territories of existing EISs: data uncertainty, temporal system dynamic, redundant data streams.
no code implementations • 6 May 2017 • Mahardhika Pratama, Eric Dimla, Chow Yin Lai, Edwin Lughofer
The learning process consists of three phases: what to learn, how to learn, when to learn and makes use of a generalized recurrent network structure as a cognitive component.
no code implementations • 10 Apr 2017 • Mahardhika Pratama, Plamen P. Angelov, Edwin Lughofer
The theory of random vector functional link network (RVFLN) has provided a breakthrough in the design of neural networks (NNs) since it conveys solid theoretical justification of randomized learning.
no code implementations • 1 Sep 2016 • Rajasekar Venkatesan, Meng Joo Er, Mihika Dave, Mahardhika Pratama, Shiqian Wu
In this paper, a high-speed online neural network classifier based on extreme learning machines for multi-label classification is proposed.
no code implementations • 31 Aug 2016 • Rajasekar Venkatesan, Meng Joo Er, Shiqian Wu, Mahardhika Pratama
In this paper, a novel extreme learning machine based online multi-label classifier for real-time data streams is proposed.