A Computational Model to Predict Consumer Behaviour During COVID-19 Pandemic

Susceptibility of AutoML mortality prediction algorithms to model drift caused by the COVID pandemic

BACKGROUND

Concept drift and covariate shift lead to a degradation of machine learning (ML) models. The objective of our study was to characterize sudden data drift as caused by the COVID pandemic. Furthermore, we investigated the suitability of certain methods in model training to prevent model degradation caused by data drift.

METHODS

We trained different ML models with the H2O AutoML method on a dataset comprising 102,666 cases of surgical patients collected in the years 2014-2019 to predict postoperative mortality using preoperatively available data. Models applied were Generalized Linear Model with regularization, Default Random Forest, Gradient Boosting Machine, eXtreme Gradient Boosting, Deep Learning and Stacked Ensembles comprising all base models. Further, we modified the original models by applying three different methods when training on the original pre-pandemic dataset: (Rahmani K, et al, Int J Med Inform 173:104930, 2023) we weighted older data weaker, (Morger A, et al, Sci Rep 12:7244, 2022) used only the most recent data for model training and (Dilmegani C, 2023) performed a z-transformation of the numerical input parameters. Afterwards, we tested model performance on a pre-pandemic and an in-pandemic data set not used in the training process, and analysed common features.

RESULTS

The models produced showed excellent areas under receiver-operating characteristic and acceptable precision-recall curves when tested on a dataset from January-March 2020, but significant degradation when tested on a dataset collected in the first wave of the COVID pandemic from April-May 2020. When comparing the probability distributions of the input parameters, significant differences between pre-pandemic and in-pandemic data were found. The endpoint of our models, in-hospital mortality after surgery, did not differ significantly between pre- and in-pandemic data and was about 1% in each case. However, the models varied considerably in the composition of their input parameters. None of our applied modifications prevented a loss of performance, although very different models emerged from it, using a large variety of parameters.

CONCLUSIONS

Our results show that none of our tested easy-to-implement measures in model training can prevent deterioration in the case of sudden external events. Therefore, we conclude that, in the presence of concept drift and covariate shift, close monitoring and critical review of model predictions are necessary.

Predictive modeling of consumer purchase behavior on social media: Integrating theory of planned behavior and machine learning for actionable insights

In recent times, it has been observed that social media exerts a favorable influence on consumer purchasing behavior. Many organizations are adopting the utilization of social media platforms as a means to promote products and services. Hence, it is crucial for enterprises to understand the consumer buying behavior in order to thrive. This article presents a novel approach that combines the theory of planned behavior (TPB) with machine learning techniques to develop accurate predictive models for consumer purchase behavior. This study examines three distinct factors of the theory of planned behavior (attitude, social norm, and perceived behavioral control) that provide insights into the primary determinants influencing online purchasing behavior. A total of eight machine learning algorithms, namely K-nearest neighbor, Decision Tree, Random Forest, Logistic Regression, Naive Bayes, Support Vector Machine, AdaBoost, and Gradient Boosting, were utilized in order to forecast consumer purchasing behavior. Empirical findings indicate that gradient boosting demonstrates superior performance in predicting customer buying behavior, with an accuracy rate of 0.91 and a macro F1 score of 0.91. This holds true when all factors, namely attitude (ATTD), social norm (SN), and perceived behavioral control (PBC), are included in the analysis. Furthermore, we incorporated Explainable AI (XAI), specifically LIME (Local Interpretable Model-Agnostic Explanations), to elucidate how the best machine learning model (i.e. gradient boosting) makes its prediction. The findings indicate that LIME has demonstrated a high level of confidence in accurately predicting the influence of low and high behavior. The outcome presented in this article has several implications. For instance, this article presents a novel way to combine the theory of planned behavior with machine learning techniques in order to predict consumer purchase behavior. This integration allows for a comprehensive analysis of factors influencing online purchasing decisions. Also, the incorporation of Explainable AI enhances the transparency and interpretability of the model. This feature is valuable for organizations seeking insights into factors driving predictions and the reasons behind certain outcomes. Moreover, these observations have the potential to offer valuable insights for businesses in customizing their marketing strategies to align with these influential factors.

Turkish Stock Market from Pandemic to Russian Invasion, Evidence from Developed Machine Learning Algorithm

In recent time, the two significant events; Coronavirus epidemic and Russian invasion are effecting all over the world in various aspects; healthily, economically, environmentally, and socially, etc. The first event has brought uncertainties to the economic situation in most countries based on the epidemic transmission. In addition to that, on 24th February 2022 the Russian invasion of Ukraine affected negatively almost all stock markets all over the world, but the effects are heterogeneous across countries according to their economic-political relationship or neighbourhood, etc. Due to that, the stock market price in Turkey has been affected dramatically over that period. This empirical study is the first attempts to explore the impact of Coronavirus epidemic and Russian invasion on the stock market index XU100 in Turkey by applying the developed statistical method namely elastic-net regression based on empirical mode decomposition which can precisely tackle the nonstationary and nonlinearity data. Then we performed the robustness check by applying a nonlinear techniques Markov switching regression. The data are collected from the beginning of the epidemic in Turkey from March 11, 2020 until May 31, 2022. The finding reveals that there is significant effect of the Coronavirus spreading on the Turkish stock market index, particularly during the first wave. Then after the Russian Invasion the XU100 index is effected more negatively. As the credit default swap and TL reference interest rate have a negative impact but the foreigner exchange rate has a positive significant impact on the XU100 index, and it varies according to the period of short term and long term. Moreover, the results obtained by using the robustness check shows a robust and consistent finding. In conclusion, understanding the impact of Coronavirus pandemic and Russian invasion on the Turkish stock market can provide important implications for investors, financial sectors, and policymakers.

How Banks Were Chosen and Rated in Hungary before and during the COVID-19 Pandemic

Banks can be chosen based on multiple factors, such as location, security, and e-banking functions. The characteristics of customers such as gender and age can also affect this decision. Since the digitalization of banking sped up due to the COVID-19 pandemic, the factors that affect this decision may change as well. To assess this, a questionnaire was completed by 156 respondents, and the results were evaluated using Pearson’s correlation test. According to the results, personal visits to the banks declined after the COVID-19 pandemic started. Furthermore, the number of e-bankers rose. When choosing banks, no gender-related relationships were found based on location, while older people chose different banks than their younger counterparts. The security of internet banking functions was not associated with bank choice, while the security of the mobile banking application was. Regarding the ratings of banks, males and females did not rate banks differently, and younger people tended to be more critical in their ratings. Security, accessible location, and good customer service can lead to more positive ratings as well. The findings can be used by banks in Hungary to improve their services in order to attract customers and increase their satisfaction.

Is Covid‐19 changing sustainable consumer behavior? A survey of Italian consumers

Since the beginning of 2020, the world has been hit by the SARS‐CoV‐2 virus that causes Covid‐19. To hamper its spread, policymakers of many countries have put in place strong countermeasures, including lockdowns, that have led to significant changes in people's lifestyles and daily routines. This article aims at assessing the changes caused by Covid‐19 in sustainable consumer behavior under multiple perspectives, contributing to advance knowledge at the intersection between consumer dynamics and sustainable consumer behavior literature. A survey was conducted on 1.535 Italian consumers between December 2020 and February 2021. Respondents were asked to assess the extent to which their consumption behavior—purchase frequency, willingness to pay a premium price, sense of moral duty to purchase, social influence to purchase—related to several categories of sustainable products changed due to the pandemic, as well as the extent to which the pandemic impacted on many other aspects, including their environmental awareness, concern, and habits. Results show that Covid‐19 generated relevant changes. Consumers have increased their purchase frequency and willingness to pay for sustainable products, show growing attention to environmental issues, and behave more sustainably. Further, the extent of change is strongly affected by socio‐demographic variables, such as gender, age, income, and education. For instance, women reported a higher shift towards sustainable consumption and behavior than men. Understanding these changes is important to guide marketers and policymakers to respond promptly and effectively to them and to leverage on them to foster a transition towards a more sustainable society.

COVID-19, consumer behavior, technology, and society: A literature review and bibliometric analysis

The COVID-19 crisis is among the most disruptive events in recent decades. Its profound consequences have garnered the interest of many studies in various disciplines, including consumer behavior, thereby warranting an effort to review and systematize the literature. Thus, this study systematizes the knowledge generated by 70 COVID-19 and consumer behavior studies in the Scopus database. It employs descriptive analysis, highlighting the importance of using quantitative methods and China and the US as research settings. Co-occurrence analysis further identified various thematic clusters among the studies. The input-process-output consumer behavior model guided the systematic review, covering several psychological characteristics and consumer behaviors. Accordingly, measures adopted by governments, technology, and social media stand out as external factors. However, revised marketing strategies have been oriented toward counteracting various consumer risks. Hence, given that technological and digital formats mark consumer behavior, firms must incorporate digital transformations in their process.

Recognition and the development potential of mobile shopping of customized cosmetic on untact coronavirus disease 2019 period: Focused on 40's to 60's women in Seoul, Republic of Korea

BACKGROUND

Our research results will be helpful in the development of the K-beauty customized cosmetics industry as well as consumers who are having difficulty in purchasing cosmetics due to the rapid transition to a non-face-to-face society due to COVID-19.

OBJECTIVES

This paper attempted to investigate the recognition and development potential of customized cosmetics, focusing on mobile shopping in the era of COVID-19 untact period.

METHODS

The women included in the study consumer Seoul residence in the mobile shopping experience 40-60 targets for women 380 were the people. Statistical processing of data collected by the data analysis method is analyzed using the SPSS (Statistical Package for Social Science) WIN25.0 statistical package program through the process of data coding and data cleaning.

RESULTS

As the untact era enters, the frequency of using non-face-to-face mobile shopping for customized cosmetics is increasing, and it is believed to be deeply related to the level of interest in skin and recognition of customized cosmetics, such as the use of hands-on apps that incorporate new technologies of the 4th industrial revolution. Looking at the Kaiser-Meyer-Olkin to confirm the normality and validity of the population proved its validity. January 2020 (after COVID-19 spread), the use of mobile shopping was increased. In the future, it was significantly higher to continue shopping for cosmetics using mobile devices p < 0.001.

CONCLUSION

The present study focused on the recognition and development potential of customized cosmetics on mobile shopping in the era of COVID-19 untact period. Our results suggested that the possibility of developing customized cosmetics through mobile shopping in the untact era after COVID-19 will be endless, and it is believed that various marketing strategies will be supported in the future.