Machine Learning to Predict the Adsorption Capacity of Microplastics

 Nanomaterials 2023, 13(6), 1061

Nowadays, there is an extensive production and use of plastic materials for different industrial activities. These plastics, either from their primary production sources or through their own degradation processes, can contaminate ecosystems with micro- and nanoplastics. Once in the aquatic environment, these microplastics can be the basis for the adsorption of chemical pollutants, favoring that these chemical pollutants disperse more quickly in the environment and can affect living beings. Due to the lack of information on adsorption, three machine learning models (random forest, support vector machine, and artificial neural network) were developed to predict different microplastic/water partition coefficients (log Kd) using two different approximations (based on the number of input variables). The best-selected machine learning models present, in general, correlation coefficients above 0.92 in the query phase, which indicates that these types of models could be used for the rapid estimation of the absorption of organic contaminants on microplastics.

Global Solar Irradiation Modelling and Prediction Using Machine Learning Models for Their Potential Use in Renewable Energy Applications

 Mathematics 2022, 10(24), 4746

Global solar irradiation is an important variable that can be used to determine the suitability of an area to install solar systems; nevertheless, due to the limitations of requiring measurement stations around the entire world, it can be correlated with different meteorological parameters. To confront this issue, different locations in Rias Baixas (Autonomous Community of Galicia, Spain) and combinations of parameters (month and average temperature, among others) were used to develop various machine learning models (random forest -RF-, support vector machine -SVM- and artificial neural network -ANN-). These three approaches were used to model and predict (one month ahead) monthly global solar irradiation using the data from six measurement stations. Afterwards, these models were applied to seven different measurement stations to check if the knowledge acquired could be extrapolated to other locations. In general, the ANN models offered the best results for the development and testing phases of the model, as well as for the phase of knowledge extrapolation to other locations. In this sense, the selected ANNs obtained a mean absolute percentage error (MAPE) value between 3.9 and 13.8% for the model development and an overall MAPE between 4.1 and 12.5% for the other seven locations. ANNs can be a capable tool for modelling and predicting monthly global solar irradiation in areas where data are available and for extrapolating this knowledge to nearby areas.

Modelling and Prediction of Monthly Global Irradiation Using Different Prediction Models

 Energies 2021, 14(8), 2332

Different prediction models (multiple linear regression, vector support machines, artificial neural networks and random forests) are applied to model the monthly global irradiation (MGI) from different input variables (latitude, longitude and altitude of meteorological station, month, average temperatures, among others) of different areas of Galicia (Spain). The models were trained, validated and queried using data from three stations, and each best model was checked in two independent stations. The results obtained confirmed that the best methodology is the ANN model which presents the lowest RMSE value in the validation and querying phases 1226 kJ/(m2∙day) and 1136 kJ/(m2∙day), respectively, and predict conveniently for independent stations, 2013 kJ/(m2∙day) and 2094 kJ/(m2∙day), respectively. Given the good results obtained, it is convenient to continue with the design of artificial neural networks applied to the analysis of monthly global irradiation.

Prediction Models to Control Aging Time in Red Wine

Molecules 2019, 24(5), 826

DOI: 10.3390/molecules24050826

A combination of physical-chemical analysis has been used to monitor the aging of red wines from D.O. Toro (Spain). The changes in the chemical composition of wines that occur over the aging time can be used to distinguish between wine samples collected after one, four, seven and ten months of aging. Different computational models were used to develop a good authenticity tool to certify wines. In this research, different models have been developed: Artificial Neural Network models (ANNs), Support Vector Machine (SVM) and Random Forest (RF) models. The results obtained for the ANN model developed with sigmoidal function in the output neuron and the RF model permit us to determine the aging time, with an average absolute percentage deviation below 1%, so it can be concluded that these two models have demonstrated their capacity to predict the age of wine.

Improved 1,3-propanediol production with maintained physical conditions and optimized media composition: Validation with statistical and neural approach

Biochemical Engineering Journal

https://doi.org/10.1016/j.bej.2017.07.003

This work is aimed at assessing the use of response surface methodology (RSM) and artificial neural networks (ANNs) for modelling, and predicting, the optimum parameters for 1,3-Propanediol production by Lactobacillus brevis N1E9.3.3 from glycerol and glucose co-fermentation. A preliminary study of physical parameters was conducted using Plackett-Burman design to reduce the number of input variables up to seven; i) beef extract, ii) yeast extract, iii) MgSO4·7H2O, iv) MnSO4·H2O, v) vitamin B12, vi) glycerol and vii) glucose. The traditional RSM models were improved by ANN models between a 54.08% and 12.19% in terms of root mean square error (RMSE). This study suggested that RSM and ANN can be considered as effective tools to model and predict optimum parameters for 1,3-Propanediol production by L. brevis N1E9.3.3.

A critical review on the applications of artificial neural networks in winemaking technology

Critical Reviews in Food Science and Nutrition

http://dx.doi.org/10.1080/10408398.2015.1078277

Since their development in 1943, artificial neural networks were extended into applications in many fields. Last twenty years have brought their introduction into winery, where they were applied following four basic purposes: authenticity assurance systems, electronic sensory devices, production optimization methods, and artificial vision in image treatment tools, with successful and promising results. This work reviews the most significant approaches for neural networks in winemaking technologies with the aim of producing a clear and useful review document.

Approach of Different Properties of Alkylammonium Surfactants using Artificial Intelligence and Response Surface Methodology

Tenside, Surfactants and Detergents

http://www.hanser-elibrary.com/doi/abs/10.3139/113.110483

Response surface methodology (RSM) and artificial neural networks (ANNs) architectures to predict the density, speed of sound, kinematic viscosity, and surface tension of aqueous solutions were developed. All models implemented using the root mean square error (RMSE) for training and validation phase were evaluated. The ANN models implemented show good values of R2 (upper than 0.974) and low errors in terms of average percentage deviation (APD) (lower than 2.92 %). Nevertheless, RSM models present low APD values for density and speed of sound prediction (lower than 0.31 %) and higher APD values around 5.18 % for kinematic viscosity and 14.73 % for surface tension. The results show that the different individual artificial neural networks implemented are a useful tool to predict the density, speed of sound, kinematic viscosity, and surface tension with reasonably accuracy.

Forecasting Olea Airborne Pollen Concentrations by Means of Artificial Intelligence

Fresenius Environmental Bulletin, (2015) 24, 4574-4580

A model based in an Artificial Neural Network was de- veloped in order to forecast the Olea airborne pollen concen- tration due to the allergenic power of its pollen grains. Daily data for Olea pollen and the meteorological variables docu- mented in the period 1993-2008, were used. All developed models had five input variables; i) Julian day, ii) precipita- tion, iii) humidity value, and iv) insolation hours. The model was carried out with data collected in the city of Ourense, North West of Spain. The developed model predicts the at- mospheric concentration of Olea pollen one day ahead. The model was tested with data from 2009 and 2010. The devel- oped model could be employed in allergology and health sci- ences in order to prevent effect of pollinosis. It is due the ability of obtain good predictions of threshold pollen con- centration values, which are important to take preventive measures by Health Systems. The time-lag observed in pre- diction phase may be due to the influence of other meteoro- logical parameters which have not been taken into account in the beginning, this fact is reaffirmed by studying the simi- larity between the different pollen seasons and the year 2010 (Mann-Whitney U Test with p<0.031), or due the low concen- trations of Olea pollen during the MPS (≈ 14 grains·m-3).

Influence Prediction of Alkylamines Upon Electrical Percolation of AOT-based Microemulsions Using Artificial Neural Networks

Tenside Surfactants Detergents, 52, 473-476 (2015)

doi: 10.3139/113.110399

Simulations for the electrical percolation of AOT/iC8/H2O w/o microemulsions added with alkylamines have been carried out by means of multilayer perceptron. Five variables have been elected as inputs: amine concentration, molecular weight, log P, hydrocarbon chain length (as number of carbons), and pKa. As a result, a neural model consisting in five input neurons, two middle layers (with fifteen and ten neurons respectively) and one output neuron was chosen because of its better performance, with a RMSE of 0.54 °C for the prediction set, with R2 = 0.9976.