Dragonfly Support Vector Machine Modelling of the Adsorption Phenomenon of Certain Phenols by Activated Carbon Fibres

Hentabli, Mohamed; Belhadj, Abd Elmouneïm; Dahmoune, Farid

doi:10.15255/kui.2020.073

Cited by 2 publications

(4 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The MATLAB machine learning toolbox offers the following built‐in positive kernel functions. [ 26,27 ]

Linear G ((),, x_{n}, x) goodbreak= x_{n}^{'} x

Gaussian G ((),, x_{n}, x) goodbreak= e^{\frac{- {(‖‖, x_{n -}^{'} x)}^{2}}{σ}}

Polynomial G ((),, x_{n}, x) goodbreak= {({1 + x}_{n}^{'} x)}^{n}

…”

Section: Methodsmentioning

confidence: 99%

“…Therefore, the Lagrange dual formula permits the aforementioned technique to be stretched to non‐linear functions. [ 24–26 ] A non‐linear SVM regression model is established by substituting the scalar product x 1 , x 2 with a non‐linear kernel:

G ((),, x_{1}, x_{2}) goodbreak= φ ((), x_{1}), φ ((), x_{2})

where φ ( x ) signifies a transformation that places x in a high‐dimensional space. The MATLAB machine learning toolbox offers the following built‐in positive kernel functions.…”

Section: Methodsmentioning

confidence: 99%

“…Therefore, the Lagrange dual formula permits the aforementioned technique to be stretched to non-linear functions. [24][25][26] A non-linear SVM regression model is established by substituting the scalar product x 1 , x 2 with a non-linear kernel:…”

Section: Svm For Regression Modellingmentioning

confidence: 99%

“…The MATLAB machine learning toolbox offers the following built-in positive kernel functions. [26,27] Linear…”

Section: Svm For Regression Modellingmentioning

confidence: 99%

See 3 more Smart Citations

Investigation of the candesartan cilexetil antihypertensive drug microencapsulation by PLA‐PVP K30 biodegradable polymers: Experimental optimization and release kinetics modelling

et al. 2023

Self Cite

View full text Add to dashboard Cite

Candesartan is an angiotensin receptor blocker (ARB) used to treat hypertension. However, its poor aqueous solubility and oral bioavailability have limited its therapeutic applications. In order to increase bioavailability and control the release of candesartan condensation, microspheres containing biodegradable polymers (polyvinylpyrrolidone [PVP K30] and polylactic acid [PLA]) in different ratios were prepared by the o/o solvent evaporation method using Span 80 as a surfactant. In addition, the impact of encapsulation parameters (i.e., PVP K30 and PLA concentrations) on the encapsulation ratio and release percentage was investigated by the mixed factorial design method. The release kinetics of the microspheres was simulated by combining two methods, the Dragonfly algorithm and a support vector machine (DA‐SVM). The experimental data were in good agreement with the predicted data, with a coefficient of determination close to unity and a mean square error close to zero. Fourier‐transfer infrared spectrometry (FTIR) analysis revealed the presence of condensation in all formulations without reporting distortion in the spectra. Scanning electron microscopy (SEM) confirmed the successful synthesis of microspheres, whose sizes were between 12 and 26 μm. Formulations with a PLA‐drug ratio of 6:1 (N15, N17, and N18) showed the highest encapsulation efficiency (68%, 71%, and 70%, respectively), while formulations that do contain PVP K30, such as N5, N4, and N3, showed a higher release (83%, 84%, and 89%, respectively), indicating that the agent (PVP K30) enhanced the bioavailability and release of candesartan. Overall, N3 showed a higher drug release rate at 12 h and important encapsulation efficiency, making it the optimal formulation obtained in this study.

show abstract

“…The MATLAB machine learning toolbox offers the following built‐in positive kernel functions. [ 26,27 ]

Linear G ((),, x_{n}, x) goodbreak= x_{n}^{'} x

Gaussian G ((),, x_{n}, x) goodbreak= e^{\frac{- {(‖‖, x_{n -}^{'} x)}^{2}}{σ}}

Polynomial G ((),, x_{n}, x) goodbreak= {({1 + x}_{n}^{'} x)}^{n}

…”

Section: Methodsmentioning

confidence: 99%

G ((),, x_{1}, x_{2}) goodbreak= φ ((), x_{1}), φ ((), x_{2})

where φ ( x ) signifies a transformation that places x in a high‐dimensional space. The MATLAB machine learning toolbox offers the following built‐in positive kernel functions.…”