Investigating the impact of pretreatment strategies on photocatalyst for accurate CO2RR productivity quantification: A machine learning approach

Unveiling the Role of Carbonate Radical Anions in Dust‐Driven SO2 Oxidation

“…Before we applied the ML strategy (Raman Forest) preformed in our early studies (Y. Liu et al., 2023a; Xie et al., 2023), a data set metrics comprising the 34 cases of 11 multiple factors (mass contents (%) of SiO 2 , Al 2 O 3 , ferric oxide, manganite oxide, MD_ACP, alkali metal oxide, MD_ACP/alkali metal oxide as well as RH (%), light intensity (mW cm −2 ) and CO 2 concentration (ppm), and the resulting uptake coefficients is established, where SO 2 heterogeneous oxidation over TiO 2 , one simulated mineral dust (SMD), six authentic dust and clays under diverse conditions were investigated and applied for “tree” model training and subsequent prediction of SO 2 uptake capability on the typical dust and clay particles at specific pollution condition of concern (Figure 6a). Application of the first half of data as a training data set gives a good prediction of the rest half test data set (Figure 6b, details available in ML methodology and ML workflow presented in Figure S2 in Supporting Information ).…”