The Capability Analysis on the Characteristic Selection Algorithm of Text Categorization Based on F1 Measure Value

He, Suqin; Cao, Jinshan; Guo, Ruixu; Wang, Guijun

doi:10.1109/imccc.2012.180

Cited by 8 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Two scenarios are constructed to track the triangle and asymmetric dodecagon targets, respectively. In order to test the effectiveness and robustness of the proposed method, we compare the STGP-PHD filter with its corresponding smoothing filter and the PHD based on GP model in the number of targets, the mean shape precision Pu, the mean shape recall Ru, F1-Measure value [28] and COT, where…”

Section: Performance Evaluationmentioning

confidence: 99%

“…Two scenarios are constructed to track the triangle and asymmetric dodecagon targets, respectively. In order to test the effectiveness and robustness of the proposed method, we compare the STGP‐PHD filter with its corresponding smoothing filter and the PHD based on GP model in the number of targets, the mean shape precision Pu , the mean shape recall Ru , F1‐Measure value [28] and COT, where

Pu=\frac{1}{{N}_{MC}}\sum\limits _{i=1}^{{N}_{MC}}\frac{Area\left({T}_{j}^{i}\cap {E}_{j}^{i}\right)}{Area\left({E}_{j}^{i}\right)}

Ru=\frac{1}{{N}_{MC}}\sum\limits _{i=1}^{{N}_{MC}}\frac{Area\left({T}_{j}^{i}\cap {E}_{j}^{i}\right)}{Area\left({T}_{j}^{i}\right)}

{T}_{j}^{i}

represents the real shape of the target,

{E}_{j}^{i}

represents the estimated shape of the target, Area ( a ) is the area of a , and N MC represents the number of MC experiments. (Algorithm 1).…”

Section: Simulationmentioning

confidence: 99%

See 1 more Smart Citation

A probability hypothesis density filter for tracking non‐rigid extended targets using spatiotemporal Gaussian process model

Zhou

Xie

et al. 2022

IET Signal Processing

View full text Add to dashboard Cite

This paper proposes a random finite set (RFS)‐based algorithm to deal with the tracking problem of multiple non‐rigid extended targets (MNRET) with irregular shapes in the presence of clutter, false alarms and missed detection. The extensions of targets are modelled by spatiotemporal Gaussian process, which is augmented with internal reference point (IRP) modelling the kinematics to construct the state of MNRET. The probability hypothesis density (PHD) filter is employed to propagate the first‐order moment of the RFS of MNRET. A suitable predicted likelihood of MNRET for the optimal partition is given, and the filter recursion is presented along with the necessary approximations and assumptions. More importantly, the closed‐form implementation and its corresponding smoothing filter are derived by converting the posterior density to Gaussian mixture form. Simulation results show the robustness and effectiveness of the proposed algorithm.

show abstract

Section: Performance Evaluationmentioning

confidence: 99%

“…Two scenarios are constructed to track the triangle and asymmetric dodecagon targets, respectively. In order to test the effectiveness and robustness of the proposed method, we compare the STGP‐PHD filter with its corresponding smoothing filter and the PHD based on GP model in the number of targets, the mean shape precision Pu , the mean shape recall Ru , F1‐Measure value [28] and COT, where

Pu=\frac{1}{{N}_{MC}}\sum\limits _{i=1}^{{N}_{MC}}\frac{Area\left({T}_{j}^{i}\cap {E}_{j}^{i}\right)}{Area\left({E}_{j}^{i}\right)}

Ru=\frac{1}{{N}_{MC}}\sum\limits _{i=1}^{{N}_{MC}}\frac{Area\left({T}_{j}^{i}\cap {E}_{j}^{i}\right)}{Area\left({T}_{j}^{i}\right)}

{T}_{j}^{i}

represents the real shape of the target,

{E}_{j}^{i}

represents the estimated shape of the target, Area ( a ) is the area of a , and N MC represents the number of MC experiments. (Algorithm 1).…”

Section: Simulationmentioning

confidence: 99%

A probability hypothesis density filter for tracking non‐rigid extended targets using spatiotemporal Gaussian process model

Zhou

Xie

et al. 2022

IET Signal Processing

View full text Add to dashboard Cite

show abstract

“…As F-score takes both recall and precision into consideration, these metrics were not used to evaluate the algorithms separately [33]. F-score is deemed more appropriate to evaluate the performance of the classifications as it presents the weighted average of precision and recall [34], [35] Table 2 depicts the results of the sentiment analysis for the English tweets.…”

Section: 𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 = 𝑇𝑃 + 𝑇𝑁 𝑇𝑃 + 𝑇𝑁 + 𝐹𝑃 + 𝐹𝑁mentioning

confidence: 99%

Social Media Analytics Using Sentiment and Content Analyses on the 2018 Malaysia’s General Election

Balakrishnan

Kaity

Rahim

et al. 2021

MJCS

View full text Add to dashboard Cite

This study analysed the political use of Twitter during the 2018 Malaysian General Election (GE14), using sentiment and content analyses to examine the patterns in online communication among urban Malaysians. Specifically, Naive Bayes, Support Vector Machine and Random Forest were used for sentiment analysis for the English tweets, with the results compared against two vectorization approaches. Content analysis involving human experts was used for the Malay tweets. Top trending hashtags were used to fetch tweets from April 15, 2018 to May 14, 2018, resulting in a curated corpus of 190 224 tweets. Naïve Bayes used along with Word2Vec outperformed all the other models with an accuracy of 63.7%, 66.8% and 64.9% for pre-GE14, GE14 and post-GE14, respectively. Generally, results indicate the majority of the sentiments to be positive in nature, followed by negative and neutral during pre-GE14, GE14 day and post-GE14 for the English speakers. Though similar sentiments were observed for the Malay speakers, the majority of their sentiments on election day were negative (i.e. 42%) as opposed to the English speakers (i.e. 31%).

show abstract

“…Normally, measure is applied as a better measure of accuracy, defined in Eq. 3 with (Shaojun et al, 2012):…”

Section: For Driver Behaviour Modellingmentioning

confidence: 99%

A Reliability Inspired Strategy for Intelligent Performance Management With Predictive Driver Behaviour: A Case Study for a Diesel Particulate Filter

et al. 2021

View full text Add to dashboard Cite

The increase availability of operational data from the fleets of cars in the field offers opportunities to deploy machine learning to identify patterns of driver behaviour. This provides contextual intelligence insight that can be used to design strategies for online optimisation of the vehicle performance, including compliance with stringent legislation. This paper illustrates this approach with a case study for a Diesel Particulate Filter, where machine learning deployed to real world automotive data is used in conjunction with a reliability inspired performance modelling paradigm to design a strategy to enhance operational performance based on predictive driver behaviour. The model-in-the-loop simulation of the proposed strategy on a fleet of vehicles showed significant improvement compared to the base strategy, demonstrating the value of the approach.

show abstract

The Capability Analysis on the Characteristic Selection Algorithm of Text Categorization Based on F1 Measure Value

Cited by 8 publications

References 9 publications

A probability hypothesis density filter for tracking non‐rigid extended targets using spatiotemporal Gaussian process model

A probability hypothesis density filter for tracking non‐rigid extended targets using spatiotemporal Gaussian process model

Social Media Analytics Using Sentiment and Content Analyses on the 2018 Malaysia’s General Election

A Reliability Inspired Strategy for Intelligent Performance Management With Predictive Driver Behaviour: A Case Study for a Diesel Particulate Filter

Contact Info

Product

Resources

About