Temporal sampling forest ( $$\varvec{\textit{TS-F}}$$ TS - F ): an ensemble temporal learner

Ooi, Shih Yin; Tan, Shing Chiang; Cheah, Wooi Ping

doi:10.1007/s00500-016-2242-7

Cited by 8 publications

(6 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To correctly predict future events over temporal data, the existing temporal value should not be ignored. Several ensemble learning techniques are extended to consider the temporal value during machine learning such as temporal sampling forest [15], Bagged.ETS.MBB [16], and time series forest [17]. Table 2 points out their details.…”

Section: Related Workmentioning

confidence: 99%

“…Table 2 points out their details. While some of the previous studies were implemented using the ensembles of exponential smoothing methods [16], some of them were implemented by developing a temporal sampling module in the attribute level [15]. New methods based on the specialties of random forest were introduced in some of the studies [15,17].…”

Section: Related Workmentioning

confidence: 99%

“…While some of the previous studies were implemented using the ensembles of exponential smoothing methods [16], some of them were implemented by developing a temporal sampling module in the attribute level [15]. New methods based on the specialties of random forest were introduced in some of the studies [15,17]. Unlike previous studies, we propose a new method based on the rationale behind bagging by modifying the sample selection stage by giving weights to instances in the training set according to their up-to-dateness.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Temporal bagging: a new method for time-based ensemble learning

TÜYSÜZOĞLU,

BİRANT,

KIRANOĞLU

2022

Turkish Journal of Electrical Engineering and Computer Sciences

View full text Add to dashboard Cite

One of the main problems associated with the bagging technique in ensemble learning is its random sample selection in which all samples are treated with the same chance of being selected. However, in time-varying dynamic systems, the samples in the training set have not equal importance, where the recent samples contain more useful and accurate information than the former ones. To overcome this problem, this paper proposes a new time-based ensemble learning method, called temporal bagging (T-Bagging). The significant advantage of our method is that it assigns larger weights to more recent samples with respect to older ones, so it reduces the selection chances of former samples, and, thus, it addresses the adaptation to changes in dynamic systems. The experiments show that the proposed T-Bagging method improves the prediction accuracy of the model compared to the standard bagging method on temporal data.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Temporal bagging: a new method for time-based ensemble learning

TÜYSÜZOĞLU,

BİRANT,

KIRANOĞLU

2022

Turkish Journal of Electrical Engineering and Computer Sciences

View full text Add to dashboard Cite

show abstract

“…Although it has been proven as a powerful technique, the standard RF procedure has some limitations. To overcome these limitations, different variants and extensions of RF have been proposed recently such as causal forest (Wager & Athey, 2018), modified random forest (MRF; An & Suh, 2020), recurring concepts adaptive random forests (RCARF; Suarez-Cetrulo et al, 2019), syncretic cost-sensitive random forest (SCSRF; Rao et al, 2020), random M5 model forest (RM5MF; Samat et al, 2018), oblique random forest (obRaF; Katuwal et al, 2020), twin-bounded support vector machine-based random forest (TBRaF; Ganaie et al, 2020), random forest with Gaussian process (RF-GP; Wu et al, 2020), ensemblemargin based random forest (EMRF; Feng et al, 2019), Sigma-z RF (Fornaser et al, 2018), random credal random forest (RCRF; Abellan et al, 2018), quantile forests (QF; Duroux & Scornet, 2018), temporal sampling forest (TS-F; Ooi et al, 2017), static-dynamic (SD) forest (Jang et al, 2017), variable importance-weighted random forests (viRF; Liu & Zhao, 2017), adaptive random forest (ARF; Gomes et al, 2017), genetic algorithm optimized random forest (RFGA; Naghibi et al, 2017), improved random forest (IRF; Zhou & Wang, 2012), sequential forward selection random forest (SFS-RF) and sequential backward selection random forest (SBS-RF; Bernard et al, 2009).…”

Section: Related Workmentioning

confidence: 99%

Multi‐view rank‐based random forest: A new algorithm for prediction in eSports

Birant

2021

Expert Systems

View full text Add to dashboard Cite

The main problem associated with the random forest (RF) algorithm is its application of random feature subset selection technique over a single vector. In this technique, the irrelevant or redundant features in a single-view data have equal chances with the important features to be selected for training a classifier, leading to misclassification. To overcome this problem, this article proposes a novel algorithm, called multi-view rank-based random forest (MVRRF). The proposed algorithm builds a set of decision trees from multi-view data by using a rank-based feature selection strategy. The main advantages of our method are that (i) it extends the RF algorithm for multi-view learning, and (ii) it reduces the chances of irrelevant and redundant feature selection, and thus it usually improves the accuracy, generalizability and robustness of the classification models. The aim of our study is to predict the match result of the game League of Legends in electronic sports (eSports). Thus, the eSports teams can define trustworthy strategies through important features. The proposed method can be successfully used in other fields as well as eSports. The experiments that were conducted on a publicly available eSports dataset show that the proposed MVRRF algorithm (93.32%) outperforms the standard RF algorithm (86.38%) on multi-view data in terms of accuracy. Furthermore, the experimental results also show that our method achieved higher performance than the methods tested in the state of the art studies on the same dataset.

show abstract

“…For many multi-class recognition tasks, RF has shown its effectiveness [30,31]. The RF consists of a number of decision trees.…”

Section: Two-layer Random Forestmentioning

confidence: 99%

Simultaneous Indoor Tracking and Activity Recognition Using Pyroelectric Infrared Sensors

Luo

Guan

Tan

et al. 2017

Sensors

View full text Add to dashboard Cite

Indoor human tracking and activity recognition are fundamental yet coherent problems for ambient assistive living. In this paper, we propose a method to address these two critical issues simultaneously. We construct a wireless sensor network (WSN), and the sensor nodes within WSN consist of pyroelectric infrared (PIR) sensor arrays. To capture the tempo-spatial information of the human target, the field of view (FOV) of each PIR sensor is modulated by masks. A modified partial filter algorithm is utilized to decode the location of the human target. To exploit the synergy between the location and activity, we design a two-layer random forest (RF) classifier. The initial activity recognition result of the first layer is refined by the second layer RF by incorporating various effective features. We conducted experiments in a mock apartment. The mean localization error of our system is about 0.85 m. For five kinds of daily activities, the mean accuracy for 10-fold cross-validation is above 92%. The encouraging results indicate the effectiveness of our system.

show abstract

Temporal sampling forest ( $$\varvec{\textit{TS-F}}$$ TS - F ): an ensemble temporal learner

Cited by 8 publications

References 39 publications

Temporal bagging: a new method for time-based ensemble learning

Temporal bagging: a new method for time-based ensemble learning

Multi‐view rank‐based random forest: A new algorithm for prediction in eSports

Simultaneous Indoor Tracking and Activity Recognition Using Pyroelectric Infrared Sensors

Contact Info

Product

Resources

About