Weighted
            <i>k</i>
            ‐nearest neighbour model for indoor VLC positioning

Tuấn, Nguyễn Văn; Minh, Hoa Le; Burton, Andrew

doi:10.1049/iet-com.2016.0961

Cited by 41 publications

(36 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For datasets that require a quick prediction kNN can be the best choice. As seen in Table III using the optimal number of k effects the % accuracy is greatly [22]. When using the low fixed odd value there is high chance of overfitting and using the high fixed odd value less than √ (n is number of data instances) does not guarantee better performance.…”

Section: Discussionmentioning

confidence: 99%

Impact of Different Data Types on Classifier Performance of Random Forest, Naïve Bayes, and K-Nearest Neighbors Algorithms

Singh¹,

Halgamuge²,

Lakshmiganthan³

2017

ijacsa

View full text Add to dashboard Cite

Abstract-This study aims to evaluate impact of three different data types (Text only, Numeric Only and Text + Numeric) on classifier performance (Random Forest, k-Nearest Neighbor (kNN) and Naïve Bayes (NB) algorithms). The classification problems in this study are explored in terms of mean accuracy and the effects of varying algorithm parameters over different types of datasets. This content analysis has been examined through eight different datasets taken from UCI to train models for all three algorithms. The results obtained from this study clearly show that RF and kNN outperform NB. Furthermore, kNN and RF perform relatively the same in terms of mean accuracy nonetheless kNN takes less time to train a model. The changing numbers of attributes in datasets have no effect on Random Forest, whereas Naïve Bayes mean accuracy fluctuates up and down that leads to a lower mean accuracy, whereas, kNN mean accuracy increases and ends with higher accuracy. Additionally, changing number of trees has no significant effects on mean accuracy of the Random forest, however, the time to train the model has increased greatly. Random Forest and k-Nearest Neighbor are proved to be the best classifiers for any type of dataset. Thus, Naïve Bayes can outperform other two algorithms if the feature variables are in a problem space and are independent. Besides Random forests, it takes highest computational time and Naïve Bayes takes lowest. The k-Nearest Neighbor requires finding an optimal number of k for improved performance at the cost of computation time. Similarly, changing the number of attributes that effect Naïve Bayes and k-Nearest Neighbor performance nevertheless not the Random forest. This study can be extended by researchers who use the parametric method to analyze results.

show abstract

Section: Discussionmentioning

confidence: 99%

Impact of Different Data Types on Classifier Performance of Random Forest, Naïve Bayes, and K-Nearest Neighbors Algorithms

Singh¹,

Halgamuge²,

Lakshmiganthan³

2017

ijacsa

View full text Add to dashboard Cite

show abstract

“…4(a) shows the sampling points of Detector 1), which is used as input to the IPWRL algorithm. For comparison, the above test samples are also the input into the conventional RSS algorithm [10], the PWRL algorithm [20], and the KNN algorithm [13]. In order to collect training data for the KNN algorithm, we take 49 points at the same height for three times, 25 of which coincide with the points corresponding to test samples (i.e., 25 red points shown in Fig.…”

Section: Experiments Setupmentioning

confidence: 99%

“…It has been found that a receiver based on multiple detectors outperforms the one based on a single detector in terms of inter-cell interference mitigation [11]. Moreover, to improve the positioning accuracy, machine learning (ML) algorithms, especially supervised learning (SL), have been introduced to the VLP [12], such as k-nearest neighbors (KNN) [13], back-propagation [14], random forest based classifiers and adaBoost based classifiers [15]. However, performance of the SL assisted VLP systems is largely affected by the training data.…”

Section: Introductionmentioning

confidence: 99%

“…Experimental investigations are conducted in a VLP system that is able to measure RSS to evaluate the positioning performance of the IPWRL. A comparison among the proposed IPWRL, the conventional RSS [10], the KNN [13], and the PWRL [20], is carried out. Our results reveal that the IPWRL is able to maintain the low positioning errors even when a large height difference is introduced, showing excellent robustness against deterministic errors, while the positioning errors of the other methods increase sharply.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Iterative point-wise reinforcement learning for highly accurate indoor visible light positioning

Zhang

Zhu

et al. 2019

Opt. Express

View full text Add to dashboard Cite

Iterative point-wise reinforcement learning (IPWRL) is proposed for highly accurate indoor visible light positioning (VLP). By properly updating the height information in an iterative fashion, the IPWRL not only effectively mitigates the impact of nondeterministic noise but also exhibits excellent tolerance to deterministic errors caused by the inaccurate a priori height information. The principle of the IPWRL is explained, and the performance of the IPWRL is experimentally evaluated in a received signal strength (RSS) based VLP system and compared with other positioning algorithms, including the conventional RSS algorithm, the k-nearest neighbors (KNN) algorithm and the PWRL algorithm where iterations exclude. Unlike the supervised machine learning method, e.g., the KNN, whose performance is highly dependent on the training process, the proposed IPWRL does not require training and demonstrates robust positioning performance for the entire tested area. Experimental results also show that when a large height information mismatch occurs, the IPWRL is able to first correct the height information and then offers robust positioning results with a rather low positioning error, while the positioning errors caused by the other algorithms are significantly higher.

show abstract

“…Currently, the most widely adopted fingerprint localization algorithms are theK-nearest neighbor (KNN) [19] algorithm and weighted KNN (WKNN) [20] algorithm due to the low complexity suitable for practical application. Taking an example of KNN, it uses the on-line RSS to search for K smallest Euclidean distance of known locations from the fingerprint database by root mean square errors principle.…”

Section: Related Workmentioning

confidence: 99%

Detection of an unknown radio transmitter using an enhanced K-nearest neighbor algorithm based on virtual reference point and RSSD information

Zhang

Jiang

2019

J Wireless Com Network

View full text Add to dashboard Cite

Accurate detection of the unknown radio transmitter (URT) is crucial to combat illegal occupation of radio signal resources and protect communication system from harmful signal interference. The fingerprint positioning technique based on received signal strength (RSS) is famous for requiring no extra equipment, antenna arrays, and time synchronization. However, conventional RSS-based fingerprint positioning techniques that utilize K-nearest neighbor (KNN) method are confronted with problems when the positioning target is radio transmitter with unknown emission strength and frequency. Moreover, they not only cannot realize the precise localization of the URT but also only rely on pre-set reference points in the fingerprint database. In this paper, a new KNN-based geo-location approach using received signal strength difference (RSSD) information and virtual reference point is proposed to estimate an URT location. To obtain more accurate RSSD measurements, a RSSD-based filtering method by calculating the Euclidean distance between each sampling RSSD and the average value is devised to establish the fingerprint database. To achieve higher positioning accuracy, we combine KNN technique with the virtual reference (VR) point to propose RSSD-VRKNN algorithm. The simulation results show that the proposed scheme can obtain the best positioning performance compared with the conventional KNN and weighted K-nearest neighbor (WKNN) techniques. The performance and feasibility of our proposed algorithm are verified through extensive experiments.

show abstract

Weighted k ‐nearest neighbour model for indoor VLC positioning

Cited by 41 publications

References 17 publications

Impact of Different Data Types on Classifier Performance of Random Forest, Naïve Bayes, and K-Nearest Neighbors Algorithms

Impact of Different Data Types on Classifier Performance of Random Forest, Naïve Bayes, and K-Nearest Neighbors Algorithms

Iterative point-wise reinforcement learning for highly accurate indoor visible light positioning

Detection of an unknown radio transmitter using an enhanced K-nearest neighbor algorithm based on virtual reference point and RSSD information

Contact Info

Product

Resources

About