Finding Hidden Links among Variables in a Large-Scale 4G Mobile Traffic Network Dataset Using Machine Learning

Diouf, Ndolane; Ndong, Massa; Diop, Dialo; Talla, Kharouna; Aboubaker, Beye Chedikh; Gueye, Ibrahima

doi:10.1109/iscmi53840.2021.9654806

Cited by 4 publications

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They can be used for both classification and regression. DTs have the ability to handle heterogeneous data, including ordered variables, categorical variables, or a mixture of both, according to (Ndolane et al, 2021). The maximum depth of the tree is a significant hyperparameter that can improve the performance of the DT algorithm.…”

Section: Related Workmentioning

confidence: 99%

“…RF is capable of modeling highly non-linear relationships and is a very powerful ensemble method that combines with a DT ensemble, according to (Hajjem et al, 2014). A set of decision trees called a forest, is generated from a new training dataset, which is a randomly sampled subset of the original training dataset, according to (Ndolane et al 2021). A subset of independent variables is randomly selected for tree division.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Research on the Link Between Economic Development Variables and the Rate of Access to Drinking Water in Rural Senegal Using Machine Learning

Coly,

Ly,

Diouf

et al. 2023

American Journal of Applied Sciences

View full text Add to dashboard Cite

After pre-processing a set of data collected from Senegalese state structures, official United Nations (UN), and world bank sites and after replacing missing values by interpolation using the mean between two values, we study the effect of macroeconomic development variables on the rate of access to drinking water in rural Senegal. To do this, we use Machine Learning (ML) techniques such as Linear Regression (LR), Decision Tree (DT), and Random Forest (RF) to identify a hidden correlation between the rate of access to drinking water and other variables. Based on the collected data, LR provides the best predictive accuracy, best RMSE (0.001), best MAE (0.011), and best R2 (96.9%) and help public development net received (ODA_NR_FBC) appears to be the most influential variable, predicting the rate of access to drinking water with greater precision than the other variables. This innovative approach can help us to better understand the factors influencing access to drinking water and to propose effective solutions.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%