A comparative study of <i>k</i>‐nearest neighbour techniques in crowd simulation

Vermeulen, Jordi L.; Hillebrand, Arne; Gerærts, Roland

doi:10.1002/cav.1775

Cited by 20 publications

(11 citation statements)

References 14 publications

(11 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The model outputs were regridded to a common grid of 1.4 × 1.4 • resolutions using the nearest neighbor interpolation technique. The nearest neighbor interpolation allowed better classification of similar close points by weighted average using data triangulation [43]. The land surfaces, particularly hilly terrains, were better interpolated due to sub-regionalization of grid points by the nearest cell center of an input grid [44].…”

Section: Methodsmentioning

confidence: 99%

Evaluation of Historical CMIP6 Model Simulations of Seasonal Mean Temperature over Pakistan during 1970–2014

et al. 2020

View full text Add to dashboard Cite

This work employed recent model outputs from coupled model intercomparison project phase six to simulate surface mean temperature during the June–July–August (JJA) and December–January–February (DJF) seasons for 1970–2014 over Pakistan. The climatic research unit (CRU TS4.03) dataset was utilized as benchmark data to analyze models’ performance. The JJA season exhibited the highest mean temperature, whilst DJF displayed the lowest mean temperature in the whole study period. The JJA monthly empirical cumulative distribution frequency (ECDF) range (26 to 28 °C) was less than that of DJF (7 to 10 °C) since JJA matched closely to CRU. The JJA and DJF seasons are warming, with higher warming trends in winters than in summers. On temporal scale, models performed better in JJA with overall low bias, low RMSE (root mean square error), and higher positive CC (correlation coefficient) values. DJF performance was undermined with higher bias and RMSE with weak positive correlation estimates. Overall, CanESM5, CESM2, CESM2-WACCM, GFDL-CM4, HadGEM-GC31-LL, MPI-ESM1-2-LR, MPI-ESM1-2-HR, and MRI-ESM-0 performed better for JJA and DJF.

show abstract

Section: Methodsmentioning

confidence: 99%

Evaluation of Historical CMIP6 Model Simulations of Seasonal Mean Temperature over Pakistan during 1970–2014

et al. 2020

View full text Add to dashboard Cite

show abstract

“…There are many machine learning algorithms which are used in image recognition, and the most popular ones in machine learning mainly include multi-layer perceptron (MLP) (Mirjalili 2015), random forests (Biau et al 2009), K nearest neighbour (KNN) (Vermeulen et al 2017), Naive Bayes (NB) (Amor et al 2004) and C4.5 decision tree (Quinlan 1996). Also, the above machine learning algorithms have been used popularly in handwritten digits recognition tasks.…”

Section: Review Of Machine Learning Methodsmentioning

confidence: 99%

Multiple classifiers fusion and CNN feature extraction for handwritten digits recognition

Zhao

Liu

2019

Granul. Comput.

View full text Add to dashboard Cite

Handwritten digits recognition has been treated as a multi-class classification problem in the machine learning context, where each of the ten digits (0-9) is viewed as a class and the machine learning task is essentially to train a classifier that can effectively discriminate the ten classes. In practice, it is very usual that the performance of a single classifier trained using a standard learning algorithm is varied on different datasets, which indicates that the same learning algorithm may train strong classifiers on some datasets but weak classifiers may be trained on other datasets. It is also possible that the same classifier shows different performance on different test sets, especially when considering the case that image instances can be highly diverse due to the different handwriting styles of different people on the same digits. To address the above issue, development of ensemble learning approaches have been very necessary to improve the overall performance and make the performance more stable on different datasets. In this paper, we propose a framework that involves CNN-based feature extraction from the MINST dataset and algebraic fusion of multiple classifiers trained on different feature sets, which are prepared through feature selection applied to the original feature set extracted using CNN. The experimental results show that the classifiers fusion can achieve the classification accuracy of ≥ 98%.

show abstract

“…We use grid‐based spatial partitioning, a method that in most, but the densest of scenarios significantly speeds up the process of searching for neighbours . More precisely, we place a regular grid over the living area and keep track of the agents located in each cell.…”

Section: Methodsmentioning

confidence: 99%

“…To alleviate the issues that arise from the computational complexity and increase processing speed, several approaches have been suggested. Some researchers reduced the computational workload by decreasing the update frequency (an increase in reaction time) of each individual, others focused on the optimization or change of the perception and interaction system, but as individual‐based models are very suitable for parallel execution, many researchers adapted techniques from highly parallel processing . The latter is most commonly achieved via general‐purpose computing on graphics processing units (GPGPU).…”

Section: Introductionmentioning

confidence: 99%

A hybrid model for simulating grazing herds in real time

Demšar

Blewitt

Bajec

2019

Computer Animation & Virtual

View full text Add to dashboard Cite

Computer simulations of animal groups are usually performed via individual‐based modelling, where simulated animals are designed on the level of individuals. With this approach, developers are able to capture behavioural nuances of real animals. However, modelling each individual as its own entity has the downside of having a high computational cost, meaning that individual‐based models are usually not suitable for real‐time simulations of very large groups. A common alternative approach is flow‐based modelling, where the dynamics of animal congregations are designed on the group level. This enables researchers to create real‐time simulations of massive phenomena at the cost of biological authenticity. A novel approach called hybrid modelling tries to mix the best of both worlds—precision of individual‐based models and speed of flow‐based ones. An unknown surrounding hybrid model is the question of their biological authenticity and relevance. In this study, we develop a hybrid model for the simulation of herds of grazing sheep. Through Bayesian data analysis, we show that such an approach can encompass several aspects of real‐world sheep behaviour. Our hybrid model is also extremely efficient, capable of simulating herds of more than 1,000 individuals in real time without resorting to graphics processing unit execution.

show abstract

A comparative study of k‐nearest neighbour techniques in crowd simulation

Cited by 20 publications

References 14 publications

Evaluation of Historical CMIP6 Model Simulations of Seasonal Mean Temperature over Pakistan during 1970–2014

Evaluation of Historical CMIP6 Model Simulations of Seasonal Mean Temperature over Pakistan during 1970–2014

Multiple classifiers fusion and CNN feature extraction for handwritten digits recognition

A hybrid model for simulating grazing herds in real time

Contact Info

Product

Resources

About