Statistical optimization of the phytoremediation of arsenic by Ludwigia octovalvis- in a pilot reed bed using response surface methodology (RSM) versus an artificial neural network (ANN)

In this study, the removal of arsenic (As) by plant, Ludwigia octovalvis, in a pilot reed bed was optimized. A Box-Behnken design was employed including a comparative analysis of both Response Surface Methodology (RSM) and an Artificial Neural Network (ANN) for the prediction of maximum arsenic removal. The predicted optimum condition using the desirability function of both models was 39 mg kg^?1 for the arsenic concentration in soil, an elapsed time of 42 days (the sampling day) and an aeration rate of 0.22 L/min, with the predicted values of arsenic removal by RSM and ANN being 72.6% and 71.4%, respectively. The validation of the predicted optimum point showed an actual arsenic removal of 70.6%. This was achieved with the deviation between the validation value and the predicted values being within 3.49% (RSM) and 1.87% (ANN). The performance evaluation of the RSM and ANN models showed that ANN performs better than RSM with a higher R² (0.97) close to 1.0 and very small Average Absolute Deviation (AAD) (0.02) and Root Mean Square Error (RMSE) (0.004) values close to zero. Both models were appropriate for the optimization of arsenic removal with ANN demonstrating significantly higher predictive and fitting ability than RSM.     

Artificial neural network and response surface methodology: a comparative analysis for optimizing rice straw pretreatment and saccharification

Abstract

The present study demonstrates a comparative analysis between the artificial neural network (ANN) and response surface methodology (RSM) as optimization tools for pretreatment and enzymatic hydrolysis of lignocellulosic rice straw. The efficacy for both the processes, that is, pretreatment and enzymatic hydrolysis was evaluated using correlation coefficient (R²) & mean squared error (MSE). The values of R² obtained by ANN after training, validation, and testing were 1, 0.9005, and 0.997 for pretreatment and 0.962, 0.923, and 0.9941 for enzymatic saccharification, respectively. On the other hand, the R² values obtained with RSM were 0.9965 for cellulose recovery and 0.9994 for saccharification efficiency. Thus, ANN and RSM together successfully identify the substantial process conditions for rice straw pretreatment and enzymatic saccharification. The percentage of error for ANN and RSM were 0.009 and 0.01 for cellulose recovery and for 0.004 and 0.005 for saccharification efficiency, respectively, which showed the authority of ANN in exemplifying the non-linear behavior of the system.     

Optimization of fermentation medium for triterpenoid production from <Emphasis Type="Italic">Antrodia camphorata</Emphasis> ATCC 200183 using artificial intelligence-based techniques

In this study, alteration in morphology of submergedly cultured Antrodia camphorata ATCC 200183 including arthroconidia, mycelia, external and internal structures of pellets was investigated. Two optimization models namely response surface methodology (RSM) and artificial neural network (ANN) were built to optimize the inoculum size and medium components for intracellular triterpenoid production from A. camphorata. Root mean squares error, R ², and standard error of prediction given by ANN model were 0.31%, 0.99%, and 0.63%, respectively, while RSM model gave 1.02%, 0.98%, and 2.08%, which indicated that fitness and prediction accuracy of ANN model was higher when compared to RSM model. Furthermore, using genetic algorithm (GA), the input space of ANN model was optimized, and maximum triterpenoid production of 62.84 mg l⁻¹ was obtained at the GA-optimized concentrations of arthroconidia (1.78 × 10⁵ ml⁻¹) and medium components (glucose, 25.25 g l⁻¹; peptone, 4.48 g l⁻¹; and soybean flour, 2.74 g l⁻¹). The triterpenoid production experimentally obtained using the ANN–GA designed medium was 64.79 ± 2.32 mg l⁻¹ which was in agreement with the predicted value. The same optimization process may be used to optimize many environmental and genetic factors such as temperature and agitation that can also affect the triterpenoid production from A. camphorata and to improve the production of bioactive metabolites from potent medicinal fungi by changing the fermentation parameters.     

Comparative Assessment of the Artificial Neural Network and Response Surface Modelling Efficiencies for Biohydrogen Production on Sugar Cane Molasses

This study comparatively evaluates the modelling efficiency of the Response Surface Methodology (RSM) and the Artificial Neural Network (ANN). Twenty-nine biohydrogen fermentation batches were carried out to generate the experimental data. The input parameters consisted of a concentration of molasses (50–150 g/l), pH (4–8), temperature (35–40 °C) and inoculum concentration (10–50 %). The obtained data were used to develop the RSM and ANN models. The ANN model was a committee of networks with a topology of 4-(6-10)-1 structured on multilayer perceptrons. RSM and ANN models gave R ² values of 0.75 and 0.91, respectively, with predicted optimum conditions of 150 g/l, 8 and 35 °C for molasses, pH and temperature, respectively, with differences in inoculum concentrations (10.11 and 15 %) for RSM and ANN, respectively. Upon validation, 15.12 and 119.08 % prediction errors on hydrogen volume were found for ANN and RSM, respectively. These findings suggest that ANN has greater accuracy in modelling the relationships between the considered process inputs for fermentative biohydrogen production and thus, is more reliable to navigate the optimization space.     

Artificial neural networking to estimate the leaf area for invasive plant Wedelia trilobata

Leaf area are very important parameter for the understanding of growth and physiological responses of invasive plant species under different environmental factors. This study was conducted to build non-destructive leaf area model of Wedelia trilobata that were grown in greenhouse. Regression analysis and artificial neural network (ANN) approaches were used for the development of leaf area model with the help of leaf length and width of 262 plants samples. In selection of best method under both techniques, the lower value of mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE) and higher value of R² were considered. According to the results it was found that ANN have higher value of (R² = 0.96) and lower value of error (MAE = 0.023, RMSE = 0.379, MAPE = 0.001) than regression analysis (R² = 0.94, MAE = 0.111, RMSE = 1.798, MAPE = 0.0005). It was concluded that error between predicted and actual value was less under ANN. Therefore, ANN model approach can be used as an alternating method for the estimation of leaf area. Through estimation of leaf area, invasive plant growth can predict under different environment conditions.     

Predictive modelling of Fe(III) precipitation in iron removal process for bioleaching circuits

In this study, the applicability of three modelling approaches was determined in an effort to describe complex relationships between process parameters and to predict the performance of an integrated process, which consisted of a fluidized bed bioreactor for Fe³⁺ regeneration and a gravity settler for precipitative iron removal. Self-organizing maps were used to visually evaluate the associations between variables prior to the comparison of two different modelling methods, the multiple regression modelling and artificial neural network (ANN) modelling, for predicting Fe(III) precipitation. With the ANN model, an excellent match between the predicted and measured data was obtained (R ² = 0.97). The best-fitting regression model also gave a good fit (R ² = 0.87). This study demonstrates that ANNs and regression models are robust tools for predicting iron precipitation in the integrated process and can thus be used in the management of such systems.     

Effects of oils and oil-related substrates on the synthetic activity of membrane-bound lipase from Rhizopus chinensis and optimization of the lipase fermentation media

The lipase from filamentous fungi Rhizopus chinensis, as a membrane-bound enzyme, possesses the excellent catalysis ability for esterification and transesterification reactions, and has a good potential in many industrial applications. In order to improve the synthetic activity of the lipase, the effects of oils and oil-related substrates on its production and the fermentation media optimization were investigated. Based on the results, it was suggested that oleic acid could be the important substrate for the lipase production. Among various oils and oil-related substrates, olive oil containing high content of oleic acid was the optimal one for the lipase production. Using orthogonal test and response surface methodology (RSM), the composition of fermentation media was further optimized. The optimized media for lipase synthetic activity and activity yield was composed of peptone 57.94 and 55.58 g L⁻¹, olive oil 21.94 and 22.99 g L⁻¹, maltose 12.91 and 14.34 g L⁻¹, respectively, with K₂HPO₄ 3 g L⁻¹, MgSO₄·7H₂O 5 g L⁻¹ and initial pH 6.0. Under the optimal conditions, the lipase activity and the activity yield were improved 61.5 and 93.4% comparing the results before optimization, respectively. The adequate models obtained had predicted the lipase production successfully.     

Optimization of operational conditions for adipate ester synthesis in a stirred tank reactor

Esterification of adipic acid and oleyl alcohol in a solvent-free system featuring a stirred tank reactor containing commercially immobilized Candida antarctica lipase B was performed. The process was carried out using an artificial neural network (ANN) trained by the Levenberg-Marquardt (LM) algorithm. The effects of four operative variables, temperature, time, amount of enzyme, and impeller speed, on the reaction yield were studied. By examining different ANN configurations, the best network was found to consist of seven hidden nodes using a hyperbolic tangent sigmoid transfer function. The values of the coefficient of determination (R²) and root mean squared error (RMSE) between the actual and predicted responses were determined to be 1 and 0.0058178 for training and 0.99467 and 0.622540 for the testing datasets, respectively. These results imply that the developed model was capable of predicting the esterification yield. The operative variables affected the yield, and their order of contribution was as follows: time > amount of enzyme > temperature > impeller speed. A high percentage of yield (95.7%) was obtained using a low level of enzyme (2.5% w/w), and the temperature, time, and impeller speed were 66.5°C, 354 min (about 6 h), and 500 rpm, respectively. A simple protocol for efficient substrate conversion in a solvent-free system evidenced by high enzyme stability is indicative of successful ester synthesis.     

Application of response surface methodology in medium optimization for spore production of <Emphasis Type="Italic">Coniothyrium minitans</Emphasis> in solid-state fermentation

Summary Spore production of Coniothyrium minitans was optimized by using response surface methodology (RSM), which is a powerful mathematical approach widely applied in the optimization of fermentation process. In the first step of optimization, with Plackett–Burman design, soluble starch, urea and KH²PO⁴ were found to be the important factors affecting C. minitans spore production significantly. In the second step, a 2³ full factorial central composite design and RSM were applied to determine the optimal concentration of each significant variable. A second-order polynomial was determined by the multiple regression analysis of the experimental data. The optimum values for the critical components for the maximum were obtained as follows: soluble starch 0.643 (36.43 g. l⁻¹), urea −0.544 (3.91 g l⁻¹) and KH²PO⁴ 0.049 (1.02 g l⁻¹) with a predicted value of maximum spore production of 9.94 × 10⁹ spores/g IDM. Under the optimal conditions, the practical spore production was 1.04 × 10¹⁰ spores/g IDM. The determination coefficient (R²) was 0.923, which ensure an adequate credibility of the model.     

Parameter optimization for thermostable lipase production and performance evaluation as prospective detergent additive

Abstract

Lipase based formulations has been a rising interest to laundry detergent industry for their eco-friendly property over phosphate-based counterparts and compatibility with chemical detergents ingredients. A thermo-stable Anoxybacillus sp. ARS-1 isolated from Taptapani Hotspring, India was characterized for optimum lipase production employing statistical model central composite design (CCD) under four independent variables (temperature, pH, % moisture and bio-surfactant) by solid substrate fermentation (SSF) using mustard cake. The output was utilized to find the effect of parameters and their interaction employing response surface methodology (RSM). A quadratic regression with R²?=?0.955 established the model to be statically best fitting and a predicted highest lipase production of 29.4?IU/g at an optimum temperature of 57.5?°C, pH 8.31, moisture 50% and 1.2?mg of bio-surfactant. Experimental production of 30.3?IU/g lipase at above conditions validated the fitness of model. Anoxybacillus sp. ARS-1 produced lipase was found to resist almost all chemical detergents as well as common laundry detergent, proving it to be a prospective additive for incorporation.     

River phytoplankton prediction model by Artificial Neural Network: Model performance and selection of input variables to predict time-series phytoplankton proliferations in a regulated river system

In this study, a comparison between statistical regression model and Artificial Neural Network (ANN) is given on the effectiveness of ecological model of phytoplankton dynamics in a regulated river. From the results of the study, the effectiveness of ANN over statistical method was proposed. Also feasible direction of increasing ANN models' performance was provided. A hypertrophic river data was used to develop prediction models (chlorophyll a (chl. a) 41.7 ± 56.8 μg L^− 1; n = 406). Higher time-series predictability was found from the ANN model. Failure of statistical methods would be due to the complex nature of ecological data in the regulated river ecosystems. Reduction of ANN model size by decreasing the number of input variables according to the sensitivity analysis did not have effectiveness with respect to the predictability on testing data set (RMSE of the ANN with all 27 variables, 25.7; 47.9 from using 2 highly sensitive variables; 42.9 from using 5 sensitive variables; 33.1 from using 15 variables). Even though the ANN model presented high performance in prediction accuracy, more efficient methods of selecting feasible input information are strongly requested for the prediction of freshwater ecological dynamics.     

Immobilization of Candida rugosa lipase on electrospun cellulose nanofiber membrane

A biocatalyst with high activity retention of lipase was fabricated by the covalent immobilization of Candida rugosa lipase on a cellulose nanofiber membrane. This nanofiber membrane was composed of nonwoven fibers with 200 nm nominal fiber diameter. It was prepared by electrospinning of cellulose acetate (CA) and then modified with alkaline hydrolysis to convert the nanofiber surface into regenerated cellulose (RC). The nanofiber membrane was further oxidized by NaIO₄. Aldehyde groups were simultaneously generated on the nanofiber surface for coupling with lipase. Response surface methodology (RSM) was applied to model and optimize the modification conditions, namely NaIO₄ content (2–10 mg/mL), reaction time (2–10 h), reaction temperature (25–35 °C) and reaction pH (5.5–6.5). Well-correlating models were established for the residual activity of the immobilized enzyme (R² = 0.9228 and 0.8950). We found an enzymatic activity of 29.6 U/g of the biocatalyst was obtained with optimum operational conditions. The immobilized lipase exhibited significantly higher thermal stability and durability than equivalent free enzyme.     

Statistical optimization of medium for the production of pyruvate oxidase by the recombinant Escherichia coli

Pyruvate oxidase (PyOD) is a very useful enzyme for clinical diagnostic applications and environmental monitor. Optimization of the fermentation medium for maximization of PyOD constitutively, production by Escherichia coli DH5α/pSMLPyOD was carried out. Response surface methodology (RSM) was used to optimize the medium constituents. A 2^6–2 fractional factorial design (first order model) was carried out to identify the significant effect of medium components towards PyOD production. Statistical analysis of results shows that yeast extract, ammonium sulfate and composite phosphate were significant factors on PyOD production. The optimized values of these three factors were obtained by RSM based on the result of a 2³ central composite rotatable design. Under these proposed optimized medium, the model predicted a PyOD activity of 610 U/L and via experimental rechecking the model, an activity of 670 U/L was attained.     

Alkaline lipase from a novel strain Burkholderia multivorans: Statistical medium optimization and production in a bioreactor

An alkaline lipase from Burkholderia multivorans was produced within 15 h of growth in a 14 L bioreactor. An overall 12-fold enhanced production (58 U mL⁻¹ and 36 U mg⁻¹ protein) was achieved after medium optimization following the “one-variable-at-a-time” and the statistical approaches. The optimal composition of the lipase production medium was determined to be (% w/v or v/v): KH₂PO₄ 0.1; K₂HPO₄ 0.3; NH₄Cl 0.5; MgSO₄·7H₂O 0.01; yeast extract 0.36; glucose 0.1; olive oil 3.0; CaCl₂ 0.4 mM; pH 7.0; inoculum density 3% (v/v) and incubation time 36 h in shake flasks. Lipase production was maximally influenced by olive oil/oleic acid as the inducer and yeast extract as the additive nitrogen. Plackett–Burman screening suggested catabolite repression by glucose. Amongst the divalent cations, Ca²⁺ was a positive signal while Mg²⁺ was a negative signal for lipase production. RSM predicted that incubation time, inoculum density and oil were required at their higher levels (36 h, 3% (v/v) and 3% (v/v), respectively) while glucose and yeast extract were required at their minimal levels for maximum lipase production in shake flasks. The production conditions were validated in a 14 L bioreactor where the incubation time was reduced to 15 h.     

Optimization of furfural production from hemicellulose extracted from delignified palm pressed fiber using a two-stage process

This study aims to optimize the conditions for furfural production from hemicellulose extracted from delignified palm pressed fiber (dPPF) via two-stage process: acid hydrolysis followed by dehydration, using response surface methodology (RSM). The extracted hemicellulose contained 80.8% xylose. In order to convert hemicellulose to xylose in the acid hydrolysis step, there were four important parameters consisting of reaction temperature (100–150 °C), sulfuric acid concentration (1–10% v/v), ratio of sulfuric acid to hemicellulose (L/S ratio) (10, 9, and 8 v/w), and reaction time (30–120 min). The maximum xylose production (12.58 g/L) was achieved at 125 °C, 5.5% sulfuric acid, L/S ratio of 9 mL/g for 30 min with the determination coefficient (R²) value of 0.90. For the dehydration process, two parameters; reaction temperature (120–160 °C) and reaction time (30–150 min), were optimized. The maximum furfural production (8.67 g/L) was achieved at a reaction temperature of 140 °C for 90 min with the determination coefficient (R²) value of 0.93.     

Optimization of process parameters of the Pholiota squarrosa extracellular polysaccharide by Box–Behnken statistical design

The quantitative effects of fermentation temperature, fermentation time and inoculum volume on the yield of Pholiota squarrosa extracellular polysaccharide were investigated using response surface methodology (RSM). The experimental data obtained were fitted to a second-order polynomial equation using multiple regression analysis and also analyzed by appropriate statistical methods. RSM analysis showed good correspondence between experimental and predicted values. It was found that three parameters represented significant effect. The coefficient of determination (R²) for the model was 98.5%. Probability value (P < .0001) demonstrated a very high significance for the regression model. By solving the regression equation and also by analyzing the response surface contour plots, the optimal process parameters were determined: fermentation temperature 28.57 °C, fermentation time 7.82 d and inoculum volume 12.57 ml. Under the optimal conditions the corresponding response value predicted for extracellular polysaccharide production was 853.73 μg per milliliter of fermentation liquor, which was confirmed by validation experiments.     

Phytoextraction potential of water fern (Azolla pinnata) in the removal of a hazardous dye,methyl violet 2B: Artificial neural network modelling

This study investigated the potential of Azolla pinnata (AP) in the removal of toxic methyl violet 2B (MV) dye wastewater using the phytoextraction approach with the inclusion of an Artificial Neural Network (ANN) modelling. Parameters examined included the effects of dye concentration, pH and plant dosage. The highest removal efficiency was 93% which was achieved at a plant dosage of 0.8 g (dye volume = 200 mL, initial pH = 6.0, initial dye concentration = 10 mg L^?1). A significant decrease in relative frond number (RFN), a growth rate estimator, observed at a dye concentration of 20 mg L^?1 MV indicated some toxicity, which coincided with the plant pigments studies where the chlorophyll a content was lower than the control. There were little differences in the plant pigment contents between the control and those in the presence of dye (5 to 15 mg L^?1) indicating the tolerance of AP to MV at lower concentrations. A three-layer ANN model was optimized (6 neurons in the hidden layer) and successfully predicted the phytoextraction of MV (R = 0.9989, RMSE = 0.0098). In conclusion, AP proved to be a suitable plant that could be used for the phytoextraction of MV while the ANN modelling has shown to be a reliable method for the modelling of phytoextraction of MV using AP.     

Application of Artificial Neural Network for predicting biomass growth during domestic wastewater treatment through a biological process

The biological treatment process is responsible for removing organic and inorganic matter in wastewater. This process relies heavily on microorganisms to successfully remove organic and inorganic matter. The aim of the study was to model biomass growth in the biological treatment process. Multilayer perceptron (MLP) Artificial Neural Network (ANN) algorithm was used to model biomass growth. Three metrics: coefficient of determination (R²), root mean squared error (RMSE), and mean squared error (MSE) were used to evaluate the performance of the model. Sensitivity analysis was applied to confirm variables that have a strong influence on biomass growth. The results of the study showed that MLP ANN algorithm was able to model biomass growth successfully. R² values were 0.844, 0.853, and 0.823 during training, validation, and testing phases, respectively. RMSE values were 0.7476, 1.1641, and 0.7798 during training, validation, and testing phases respectively. MSE values were 0.5589, 1.3551, and 0.6081 during training, validation, and testing phases, respectively. Sensitivity analysis results showed that temperature (47.2%) and dissolved oxygen (DO) concentration (40.2%) were the biggest drivers of biomass growth. Aeration period (4.3%), chemical oxygen demand (COD) concentration (3.2%), and oxygen uptake rate (OUR) (5.1%) contributed minimally. The biomass growth model can be applied at different wastewater treatment plants by different plant managers/operators in order to achieve optimum biomass growth. The optimum biomass growth will improve the removal of organic and inorganic matter in the biological treatment process.     

Enhanced production of amidase from <Emphasis Type="Italic">Rhodococcus erythropolis</Emphasis> MTCC 1526 by medium optimisation using a statistical experimental design

In the present work, statistical experimental methodology was used to enhance the production of amidase from Rhodococcus erythropolis MTCC 1526. R. erythropolis MTCC 1526 was selected through screening of seven strains of Rhodococcus species. The Placket–Burman screening experiments suggested that sorbitol as carbon source, yeast extract and meat peptone as nitrogen sources, and acetamide as amidase inducer are the most influential media components. The concentrations of these four media components were optimised using a face-centred design of response surface methodology (RSM). The optimum medium composition for amidase production was found to contain sorbitol (5 g/L), yeast extract (4 g/L), meat peptone (2.5 g/L), and acetamide (12.25 mM). Amidase activities before and after optimisation were 157.85 units/g dry cells and 1,086.57 units/g dry cells, respectively. Thus, use of RSM increased production of amidase from R. erythropolis MTCC 1526 by 6.88-fold.