Background: Four members of N,N-disulfo-1,1,3,3-tetramethylguanidinium chlorometallates [DSTMG]n[X], where n= 1 or 2; X= FeCl4 - , Zn2Cl6 2-, NiCl4 2-, MnCl4 2- were synthesized as solid Brønsted-Lewis acidic compounds and studied the catalytic activity with the most acidic salt for three-component synthesis of 1,2-dihydro-1-aryl-3H-naphth[1,2- e][1,3]oxazin-3-ones. Methods: N,N-disulfo-1,1,3,3-tetramethylguanidinium chlorometallates of the four transition metal cations such as Fe(III), Zn(II), Ni(II) and Mn(II) were prepared after treatment of the parent ionic liquid N,N-disulfo-1,1,3,3- tetramethylguanidinium chloride [DSTMG][Cl] with the respective metal chlorides in different mole fractions at 75 ºC. The synthesis of 1,2-dihydro-1-aryl-3H-naphth[1,2-e][1,3]oxazin-3-ones was carried out via three-component reaction of 2-naphthol, aromatic aldehydes and urea under neat condition at 90 ºC using 7 mol% of the [DSTMG][FeCl4] catalyst. Results: The characterization of synthesized chlorometallates were done using spectroscopic and other analytical techniques including thermogravimetric analysis and Hammett acidity studies. Among the four salts, the salt of Fe(III) ion was observed as the strong Brønsted acidic hydrophobic salt and thus chosen for the catalytic study. Conclusion: A new type of chlorometallates of guanidinium cation with composition [DSTMG]n[X], where X= FeCl4 - /Zn2Cl6 2-/ NiCl4 2-/ MnCl4 2- and n= 1 or 2 were developed as –SO3H functionalized solid acids with varied thermal stability (150-250 ºC) and physisorbed water (0-20%) as observed from the thermogravimetric study. From them, the most Brønsted acidic Fe(III) salt was employed as efficient recyclable heterogeneous catalyst for the one-pot synthesis of 1,2- dihydro-1-aryl-3H-naphth[1,2-e][1,3]oxazin-3-ones in neat condition.

Background: Large financial companies are perpetually creating and updating customer scoring techniques. From a risk management view, this research for the predictive accuracy of probability is of vital importance than the traditional binary result of classification, i.e., non-credible and credible customers. The customer's default payment in Taiwan is explored for the case study. Objective: The aim is to audit the comparison between the predictive accuracy of the probability of default with various techniques of statistics and machine learning. Method: In this paper, nine predictive models are compared from which the results of the six models are taken into consideration. Deep learning-based H2O, XGBoost, logistic regression, gradient boosting, naïve Bayes, logit model, and probit regression comparative analysis is performed. The software tools such as R and SAS (university edition) is employed for machine learning and statistical model evaluation. Results: Through the experimental study, we demonstrate that XGBoost performs better than other AI and ML algorithms. Conclusion: Machine learning approach such as XGBoost effectively used for credit scoring, among other data mining and statistical approaches.