Multi-factor Model Research Articles

Developing a predictive nomogram and web-based survival calculator for locally advanced hypopharyngeal cancer: A propensity score-adjusted, population-based study.

Understanding the clinical features and accurately predicting the prognosis of patients with locally advanced hypopharyngeal squamous cell carcinoma (LA-HPSCC) is important for patient centered decision-making. This study aimed to create a multi-factor nomogram predictive model and a web-based calculator to predict post-therapy survival for patients with LA-HPSCC. A retrospective cohort study analyzing Surveillance, Epidemiology, and End Results (SEER) database from 2004 to 2015 for patients diagnosed with LA-HPSCC was conducted and randomly divided into a training and a validation group (7:3 ratio). The external validation cohort included 276 patients from Sichuan Cancer Hospital, China. The Least Absolute Shrinkage and Selection Operator (LASSO)-Cox regression analysis was used to identify independent factors associated with overall survival (OS) and cancer-specific survival (CSS), and nomogram models and web-based survival calculators were constructed. Propensity score matching (PSM) was used to compare survival with different treatment options. A total of 2526 patients were included in the prognostic model. The median OS and CSS for the entire cohort were 20 (18.6-21.3) months and 24 (21.7-26.2) months, respectively. Nomogram models integrating the seven factors demonstrated high predictive accuracy for 3-year and 5-year survival. PSM found that patients who received surgery-based curative therapy had better OS and CSS than those who received radiotherapy-based treatment (median survival times: 33 months vs 18 months and 40 months vs 22 months, respectively). The nomogram model accurately predicted patient survival from LA-HPSCC. Surgery with adjuvant therapy yielded significantly better survival than definitive radiotherapy. and should be prioritized over definitive radiotherapy.

Scale transfer learning of hyperspectral prediction model of heavy metal content in maize: From laboratory to satellite

ABSTRACT For sustainable development of the environment, there has been a growing interest in using remote sensing technology to monitor the content of heavy metals in crops. Satellite hyperspectral data, while capable of providing valuable information, is often subject to environmental factors. On the other hand, laboratory spectra are collected under controlled conditions, enabling more accurate and stable models to be built. As a result, this study used the laboratory maize spectra measured using an SVC spectrometer to establish the heavy metal content prediction model and applied it to hyperspectral satellite images (Gaofen-5 and Zhuhai-1) to achieve large-scale and rapid monitoring of heavy metal content in farmland maize. However, due to the difference between the laboratory spectra (source domain) and satellite spectra (target domain), the direct transfer of laboratory models to the satellite scale has shown poor results. To address this challenge, multiple factors, including spectral feature selection, domain adaptation, regression methods, and correction models were taken into account to construct the multi-factor combined model. The research has shown that modelling after spectral feature selection can increase the accuracy of the model while reducing its complexity. Additionally, the domain adaptation method can narrow the gap between the source domain and the target domain, improving the transfer learning performance of different domains to a certain extent. Furthermore, the addition of a constructed correction model generally improves the accuracy of the transfer model. The optimal transferable model is finally obtained, which can effectively apply the laboratory heavy metal content prediction model to satellite images, and achieve large-scale and rapid monitoring of heavy metal content in farmland maize.

Unravelling the Impact of Cyclic Mechanical Stretch in Keratoconus-A Transcriptomic Profiling Study.

Biomechanical and molecular stresses may contribute to the pathogenesis of keratoconus (KC). We aimed to profile the transcriptomic changes in healthy primary human corneal (HCF) and KC-derived cells (HKC) combined with TGFβ1 treatment and cyclic mechanical stretch (CMS), mimicking the pathophysiological condition in KC. HCFs (n = 4) and HKCs (n = 4) were cultured in flexible-bottom collagen-coated 6-well plates treated with 0, 5, and 10 ng/mL of TGFβ1 with or without 15% CMS (1 cycle/s, 24 h) using a computer-controlled Flexcell FX-6000T Tension system. We used stranded total RNA-Seq to profile expression changes in 48 HCF/HKC samples (100 bp PE, 70-90 million reads per sample), followed by bioinformatics analysis using an established pipeline with Partek Flow software. A multi-factor ANOVA model, including KC, TGFβ1 treatment, and CMS, was used to identify differentially expressed genes (DEGs, |fold change| ≥ 1.5, FDR ≤ 0.1, CPM ≥ 10 in ≥1 sample) in HKCs (n = 24) vs. HCFs (n = 24) and those responsive to TGFβ1 and/or CMS. PANTHER classification system and the DAVID bioinformatics resources were used to identify significantly enriched pathways (FDR ≤ 0.05). Using multi-factorial ANOVA analyses, 479 DEGs were identified in HKCs vs. HCFs including TGFβ1 treatment and CMS as cofactors. Among these DEGs, 199 KC-altered genes were responsive to TGFβ1, thirteen were responsive to CMS, and six were responsive to TGFβ1 and CMS. Pathway analyses using PANTHER and DAVID indicated the enrichment of genes involved in numerous KC-relevant functions, including but not limited to degradation of extracellular matrix, inflammatory response, apoptotic processes, WNT signaling, collagen fibril organization, and cytoskeletal structure organization. TGFβ1-responsive KC DEGs were also enriched in these. CMS-responsive KC-altered genes such as OBSCN, CLU, HDAC5, AK4, ITGA10, and F2RL1 were identified. Some KC-altered genes, such as CLU and F2RL1, were identified to be responsive to both TGFβ1 and CMS. For the first time, our multi-factorial RNA-Seq study has identified many KC-relevant genes and pathways in HKCs with TGFβ1 treatment under CMS, suggesting a potential role of TGFβ1 and biomechanical stretch in KC development.

Other People in Career Choice Influence Career Self-Efficacy and Individual Career Self-Management

<p style="text-align: justify;">The study aims to point out direct and indirect correlations between others' roles on career self-efficacy, and career self-management. This research uses a multifactor correlational model, samples were 342 students. The research instruments are Career Self-Management Self-Efficacy Scale (CEDLE-Scale), Career Decision Self-Efficacy (CDSE Short-Form), and Role Model Influence Career Decisions Scale (RIMICDS). Data were analyzed using path analysis with AMOS 23. The findings show that 1) the variable of others' role directly influences career self-efficacy; 2) career self-management directly influences the career self-efficacy on career self-efficacy; 3) others' influence on career decision directly influences the career self-management, and 4) others' role on the career decision indirectly influences self-efficacy moderated by career self-management. Support, guidance, and inspiration from others will ease someone in deciding on a career. Guidance for managing a career will improve career self-efficacy. It will then positively contribute to the self-confidence to pursue the career.</p>

Modeling of pertinent parameters influence on the time dependent mass transfer coefficient of particulate matter under the sink effect

One of the most crucial evaluation metrics for the performance of particle sink purification technology is the time-dependent mass transfer coefficient (TDMTC). Therefore, it is very helpful for designers and developers to accurately describe the functional relationship between different influence parameters and the TDMTC. In this paper, four influence parameters (the applied voltage (V), interelectrode distance (dc), porosity (P), and shape (n) of the collecting electrode) were considered, and then non-linear multiple regression (NLMR) and multi-gene genetic programming (MGGP) methods were used to establish prediction models of the TDMTC. Results showed that V and n were the most significant factors, followed by dc and P. Both multi-factor models could accurately predict the TDMTC under the sink effect with a maximum prediction error of 20% and 15%, respectively. Moreover, for particulate matter (PM) with different size fractions, MGGP models could improve the prediction accuracy by 5–10% compared to NLMR models.

Analysis of Risky Riding Behavior Characteristics of the Related Road Traffic Injuries of Electric Bicycle Riders.

Electric bicycle (EB) riders, being vulnerable road users (VRUs), are increasingly becoming victims of road traffic injuries (RTIs). This study aimed to determine the current status and epidemiological characteristics of RTIs among EB riders through a questionnaire survey and roadside observations in Shantou to provide a scientific basis for the prevention and control of electric bicycle road traffic injuries (ERTIs). A total of 2412 EB riders were surveyed, and 34,554 cyclists were observed in the study. To analyze the relationship between riding habits and injuries among EB riders, chi-square tests and multi-factor logistic regression models were employed. The findings reveal that the prevalence of ERTIs in Shantou was 4.81%, and the most affected group was children under 16 years old, accounting for 9.84%. Risky behavior was widespread among EB riders, such as the infrequent wearing of safety helmets, carrying people on EBs, riding on sidewalks, and listening to music with headphones while bicycling. Notably, over 90% of those who wore headphones while bicycling engaged in this risky behavior. The logistic regression analysis showed that honking the horn (odds ratio (OR): 2.009, 95% CI: 1.245-3.240), riding in reverse (OR: 4.210, 95% CI: 2.631-6.737), and continuing to ride after a fault was detected (OR: 2.010, 95% CI: 1.188-3.402) all significantly increased the risk of ERTIs (all p < 0.05). Risky riding behavior was significantly less observed at traffic intersections with traffic officers than at those without (all p < 0.001).

FORECASTING THE EXCHANGE RATE OF THE UKRAINIAN HRYVNIA USING MACHINE LEARNING METHODS

This article describes the concept of currency exchange rate and the typology of various factors that influence it. A multifactor regression model was constructed to investigate the influence of factors on the exchange rate of the Ukrainian hryvnia and to forecast the dynamics of this rate based on the studied factors using Data Science technologies.&#x0D; The purpose of this work is to study the peculiarities of the formation of the exchange rate of the Ukrainian hryvnia, the characteristics of the influence of various external factors on this rate, and the creation of an effective forecasting model of the Ukrainian national currency rate, based on a certain number of fundamental financial and economic factors that influence this rate.&#x0D; Macroeconomic indicators that theoretically have an impact on the dynamics of the currency exchange rate were chosen to build the model. Data on the exchange rate of the Ukrainian hryvnia to the US dollar and economic indicators for selected factors were collected from 2010 to September 2022. During the implementation of the task, the collected data was processed, brought into a uniform form, and normalized. Machine learning methods were used for regression modeling, specifically the XGBoost gradient boosting method.&#x0D; As a result, a retrospective forecast of the Ukrainian hryvnia exchange rate was obtained, based on factor variables, and an estimate of the impact of each selected feature on the currency exchange rate was calculated. The scientific novelty of this work lies in the application of modern machine learning methods and technologies for the analysis, modeling, and forecasting of the exchange rate of the Ukrainian national currency.&#x0D; The practical significance of this article lies in the possibility of using the proposed approaches to forecasting the exchange rate of the Ukrainian hryvnia with the use of machine learning methods by all interested parties, including financial institutions of Ukraine, to achieve stability of the national currency, which in turn will affect the development of the national economy as a whole and the welfare of the population of the country.

Abnormal Returns From Hedging, Firm Size, and the Fama and French Multifactor Models

I use the Fama and French (2015) five-factor model to reexamine the seemingly anomalous result of Nelson, Moffitt, and Affleck-Graves (2005), who document significant positive abnormal returns for firms that hedge. Contrary to their results, using the five-factor model on a new sample of U.S. firms from 2013 – 2021, I observe significant negative monthly abnormal returns of -0.190% (-2.26% annually) for firms using derivative securities (hedgers). My result is consistent with poorly diversified managers engaging in costly hedging behavior that benefits management at the cost of shareholders. When I divide the sample by size (total assets), I find that the significant negative abnormal returns are confined only to large firms, offering no support for the economies of scale or managerial sophistication hypotheses.

A Comparison of Competing Asset Pricing Models: Empirical Evidence from Pakistan

In recent years, the rapid and significant development of emerging markets has globally led to insight from potential investors and academicians seeking to assess these markets in terms of risk inheritance. Therefore, this study aims to explore the validity and applicability of the capital asset pricing model (henceforth CAPM) and multi-factor models, namely Fama–French models, in Pakistan’s stock market for the period of June 2010–June 2020. This study collects data on 173 non-financial firms listed on the Pakistan stock exchange, namely the KSE-100 index, and follows Fama-MacBeth’s regression methodology for empirical estimation. The empirical findings of this study conclude that small portfolios (small-size companies) earn considerably higher returns than big portfolios (large-size companies). Ultimately, the risk associated with portfolio returns is reported to be higher for small portfolios (small-size companies) than for big portfolios (large-size companies). According to the regression output, the CAPM was found to be valid for explaining the market risk premium above the risk-free rate. Similarly, the FF three-factor model was found to be valid for explaining time-series variation in excess portfolio returns. Later, we added human capital into FF three- and five-factor models. This study found that the human capital base six-factor model outperformed the other competing asset pricing models. The findings of this study indicate that small portfolios (small-size companies) earn more returns than big portfolios (large-size companies) to reward the investor for taking extra risks. Investors may benefit by timing their investments to maximize stock returns. Company investment in human capital adds reliable information, replicates the value of the company and, in the long term, helps investors make rational decisions.

Regional differences in China's electric vehicle sales forecasting: Under supply-demand policy scenarios

Policy incentives are the key driving force for the electric vehicle (EV) market cultivation. During the EV market cultivation in different regions, the supply-side and demand-side policies have different effects. Accurately forecasting the stage characteristics and response sensitivity of EV sales under supply-demand side policy scenarios, which is crucial to the EV promotion policies design. This study selects the EV sales in 31 provinces with data available, as the basis for decision-making, and proposes a multi-factor prediction model integrating grey relation analysis (GRA), discrete wavelet transform (DWT), and bidirectional long short-term memory. Combined with the development difference of 31 provinces, the penetration of China's EV market under the benchmark, supply, demand, and ideal scenarios are verified. The experimental results show that the average Mean Absolute Percentage Error (MAPE) of the GRA-DWT-BiLSTM model is 9.884, and the 31 samples show good applicability for EV sales forecasting. In 2027, the growth rate of China's EV sales in the demand-side scenario will exceed the supply-side scenario. Under the ideal scenario, China's EV penetration rate will reach 27.31%, 42.40%, and 52.97% in 2025, 2030, and 2035 respectively. The forecast results provide a decision-making basis for China's EV market sequential supply-demand side policies.

Anatomical and topographical characteristics of the eye in the early stages of primary angle closure disease

PURPOSE. To determine the anatomical and topographic features of the anterior and posterior segments of the eye in primary angle closure (PAC) and compare them with those in the primary angle closure suspects (PACs) and in controls.METHODS. This prospective study included 120 patients aged 41 to 80 years (60 eyes — PAC, 30 — PACs, 30 eyes — without ophthalmic pathology). The studied parameters included spherical equivalent (SE), intraocular pressure (IOP), choroidal thickness in the fovea (CTf), axial length (AL), anterior chamber depth (ACD), lens volume (LV), iris curvature and its thickness, anterior chamber angle profile: AOD500, AOD750, TISA500, TISA750.RESULTS. The PAC eyes differed from the control eyes by increased SE, LV, IOP, iris curvature, CTf, reduced ACD, AL, and the anterior chamber profile (AOD500, AOD750, TISA500, TISA750) in vertical sectors (p&lt;0.01). Iris thickness was comparable between the groups.The PAC eyes differed from the PACs by increased SE, LV, IOP, reduced ACD, AL, AOD500 in the superior sector, and the profile of the anterior chamber angle in the inferior sector (p&lt;0.01). Iris thickness, its curvature, as well as the profile of the anterior chamber in the upper sector and CTf were comparable.Threshold values that distinguish the PAC eyes and the PACs were determined: LV — 0.656 mm, AOD500 — 0.131 mm, TISA500 — 0.051 mm2, TISA750 — 0.093 mm2 in the lower sector and IOP 21 mm Hg. In the multifactor model adjusted for age, gender and AL, a negative correlation of iris thickness with CTf was established both in the PAC eyes and the PACs (all p&lt;0.01), but not in the controls (all p&gt;0.01). CONCLUSION. Parameters of the PAC eyes and the PACs significantly differ from the norm. LV and the anterior chamber profile in the inferior sector are the most informative for differentiating the PAC eyes from the PACs. Negative correlation between the iris and the choroidal thickness suggests that the choroid plays a role in the pathogenesis of primary angle closure disease.

ESG Scoring System Construction: Portfolio Investment Based on Machine Learning

The majority of existing ESG rating systems in the Chinese market are based on categorical classification ratings, and as a result of the voluntary disclosure system, rating data provided by rating organizations is occasionally absent or delayed. This article employs natural language processing (NLP) to extract keywords such as green, clean, renewable, poverty alleviation, and moral from the financial reports of CSI 300 constituent companies, and then counts their corresponding frequencies in order to construct percentage ESG ratings that address the discontinuity, imprecision, and time lag inherent in the original ratings. This article employs a self-normalized neural network (SNN) to develop a multi factor model based on the suggested ESG ratings and then conducts sector neutral hierarchical back-testing to compare the proposed rating to the traditional ratings. The results indicate that the model generated using the ESG ratings developed in this research yields a higher rate of return than the model built using traditional ESG ratings, and the model constructed without an ESG factor. This may be because deriving ESG ratings directly from financial statements eliminates the risk of corporate falsification or whitewashing of accounts. This work adds to the body of knowledge by proposing a novel approach to constructing an ESG scoring system and incorporating it into portfolio investments to maximize returns.

A Test to Distinguish Monotone Homogeneity from Monotone Multifactor Models

The goodness-of-fit of the unidimensional monotone latent variable model can be assessed using the empirical conditions of nonnegative correlations (Mokken in A theory and procedure of scale-analysis, Mouton, The Hague, 1971), manifest monotonicity (Junker in Ann Stat 21:1359–1378, 1993), multivariate total positivity of order 2 (Bartolucci and Forcina in Ann Stat 28:1206–1218, 2000), and nonnegative partial correlations (Ellis in Psychometrika 79:303–316, 2014). We show that multidimensional monotone factor models with independent factors also imply these empirical conditions; therefore, the conditions are insensitive to multidimensionality. Conditional association (Rosenbaum in Psychometrika 49(3):425–435, 1984) can detect multidimensionality, but tests of it (De Gooijer and Yuan in Comput Stat Data Anal 55:34–44, 2011) are usually not feasible for realistic numbers of items. The only existing feasible test procedures that can reveal multidimensionality are Rosenbaum’s (Psychometrika 49(3):425–435, 1984) Case 2 and Case 5, which test the covariance of two items or two subtests conditionally on the unweighted sum of the other items. We improve this procedure by conditioning on a weighted sum of the other items. The weights are estimated in a training sample from a linear regression analysis. Simulations show that the Type I error rate is under control and that, for large samples, the power is higher if one dimension is more important than the other or if there is a third dimension. In small samples and with two equally important dimensions, using the unweighted sum yields greater power.

The Rise and Fall of Financial Flows in EU 15: New Evidence Using Dynamic Panels with Common Correlated Effects.

This paper assesses capital mobility for a panel of 15 European countries for the period 1970-2019 using dynamic common correlated effects modeling as proposed in Chudik and Pesaran (J Econ 188(2):393-420, 2015). In particular, we account for the existence of cross section dependence, slope heterogeneity, nonstationarity and endogeneity in a multifactor error correction model (ECM) that includes one homogeneous break. The analysis also identifies the heterogeneous structural breaks affecting the relationship for each of the individual countries. The ECM setting allows for a complete assessment of the domestic saving-investment relationship in the long-run as well as two other elements usually neglected: short-run capital mobility and the speed of adjustment. When we account for a single homogeneous break, this is found at the euro inception. We obtain that long-run capital mobility is high but not perfect yet. We also provide empirical evidence for the Ford and Horioka (Appl Econ Lett, 24(2), 95-97, 2017)'s hypothesis, who argue that goods market integration is a necessary condition to obtain zero correlation between domestic saving-investment. Our results stress the role played by the euro as a booster for both financial and real integration. However, a complete degree of economic integration has not been fully achieved. Short-run capital was highly mobile for the whole period, with some exceptions, coinciding with turmoil episodes. Additionally, from the application of the CS-DL threshold analysis proposed by Chudik et al. (Adv Econ, 36, 85-135, 2016), we find that economic risk and openness play a key role in capital mobility.

REFER: Randomized Online Factor Selection Framework for portfolio Management

Portfolio management is a critical problem in both machine learning and finance communities. To predict the returns of assets, existing studies have been leveraging side information to mine price-sensitive indicators, i.e., factors. However, with the brisk expansion of factor collection, existing factor selection methods face two main issues, i.e., high-cost and low-precision. In this paper, we first formalize the task of online factor selection as an online learning problem where a learner selects a set of factors in each round and aims to minimize the long-term regret. Then, we propose a Randomized onlinE Factor sElection fRamework, named REFER, which not only is particularly devised to address the above two issues, but also can widely serve for any existing multifactor models. Specifically, by studying the regret of existing factor-selecting policies, we propose two randomized policies along with their bandit variants that achieve sublinear regrets. The bandit variants further improve computational efficiency, and achieve a balanced trade-off between cost and precision. Finally, both theoretical analysis and extensive experiments on the real-world dataset demonstrate the effectiveness of our proposed framework and policies in terms of comprehensive evaluation criteria.

The Purification and Characterization of a Cutinase-like Enzyme with Activity on Polyethylene Terephthalate (PET) from a Newly Isolated Bacterium Stenotrophomonas maltophilia PRS8 at a Mesophilic Temperature

A polyethylene terephthalate (PET)-degrading bacterium identified as Stenotrophomonas maltophilia PRS8 was isolated from the soil of a landfill. The degradation of the PET bottle flakes and the PET prepared as a powder were assessed using live cells, an extracellular medium, or a purified cutinase-like enzyme. These treated polymers were analyzed using Fourier transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM). The depolymerization products, identified using HPLC and LC-MS, were terephthalic acid (TPA), mono(2-hydroxyethyl)-TPA (MHET), and bis(2-hydroxyethyl)-TPA (BHET). Several physicochemical factors were optimized for a better cutinase-like enzyme production by using unique single-factor and multi-factor statistical models (the Plackett–Burman design and the central composite design software). The enzyme was purified for homogeneity through column chromatography using Sephadex G-100 resin. The molecular weight of the enzyme was approximately 58 kDa. The specific activity on para nitrophenyl butyrate was estimated at 450.58 U/mg, with a purification of 6.39 times and a yield of 48.64%. The enzyme was stable at various temperatures (30–40 °C) and pH levels (8.0–10.0). The enzyme activity was significantly improved by the surfactants (Triton X-100 and Tween-40), organic solvent (formaldehyde), and metals (NiCl2 and Na2SO4). The extracellular medium containing the cutinase-type enzyme showed a depolymerization yield of the PET powder comparable to that of Idonella skaiensis IsPETase and significantly higher than that of Humicola insolens thermostable HiCut (HiC) cutinase. This study suggests that S. maltophilia PRS8 is able to degrade PET at a mesophilic temperature and could be further explored for the sustainable management of plastic waste.

The effective ways of combating monopoly in business sphere in condition of digitalization

In world practice, it is important to develop targeted strategies aimed at further improving the business environment, implementing active business projects to en sure sustainable economic development, pursuing economic policies that have a positive effect on business development, conducting research aimed at using blockchain technology as well as ensuring interest in business development by state and society. The issues of doing business after the global crisis caused by the pandemic, statistical observation, and forecasting issues, statistical assessment of the impact of the pandemic on business entities, and improving methods of statistical analysis of business development processes during and after the pandemic became important as never before. As a result of the conducted research, the authors have reached assessing quantitative and qualitative indicators of factors affecting business development processes has been developed, and software has been created that makes the accurate analyzes on the basis of blockchain of the data, a comprehensive analytical approach was developed based on statistical indicators characterizing business development trends in the country, in the SNA sectors, in the context of key industries and regions, multifactor empirical models were developed and forecast options for 2021-2026 were proposed using the scenario method. The theoretical approaches and the initial data used are taken from offcial sources, based on the data of business entities operating in Uzbekistan, proposals, and recommendations implemented in practice, statistical data of the State Committee on Statistics of the Republic of Uzbekistan, as well as primary data obtained during the study. The proposed blockchain stages in the busine process accurately evaluate the results that meet the requirements of digital economy.

ANALYSIS OF THE URBAN HEAT ISLAND USING MICROCLIMATE SIMULATION FOR URBAN QUARTER

This study demonstrates that the development of green infrastructure is an important task in the formation of ur-ban planning strategies to reduce the effect of the urban heat island and improve the ecosystem of the city. Simulation is an effective method for studying the complex mechanisms of urban climate formation at the stage of urban planning. The purpose of this study is to assess the severity of the urban heat island in relation to the quarter, taking into ac-count various scenarios of its landscaping, using modern simulation tools. Based on modeling in the ENVI-met soft-ware and computing complex of the thermal conditions of the quarter on the hottest days, its high thermal heterogenei-ty was established. The maximum temperature values are noted in roads and soils, the minimum – in green areas. The temperature conditions of the quarter changes over time. Exceeding the average temperature of the urban quarter ter-ritory over the average temperature in green areas means the possibility of forming an urban heat island. The calcula-tion established that most of the quarter is located in the urban heat island zone. This is also confirmed by an increase in air temperature in an urbanized area. Recommendations are given to mitigate the urban heat island. Based on the results of the thermal simulation of the quarter, it was established that the most effective solution compared to the orig-inal model is an increase in the area of lawn grass and shrubs by 10%, an increase in the area of trees by 12% and a decrease in asphalt pavements of paths and sites by 5.7%. Such a solution maximizes the mitigation of the urban heat island and provides a high level of comfort for the urban environment. Further research will focus on the development of a multi-factor correlation-regression model to assess the mitigation effect of urban heat islands by means of improv-ing green infrastructure.

Application of DeepForest-CQP multi-factor model in quantitative stock selection

The stock market is a chaotic system, and stock forecasting has been the research focus. This paper proposes a multi-factor model based on DeepForest-CQP to make it more applicable to the stock domain. A t-test is used for selecting factors, and orthogonalization and heteroskedasticity tests are performed for the combined factors, which are particularly important in stock forecasting. DeepForest-CQP was combined with the multi-factor model to construct a stock selection model that can achieve higher returns. The obtained multi-factor quantitative stock selection model is used to study stock selection strategies, and simulated trading is used to evaluate the multi-factor model and stock selection strategies and compare them with various machine learning multi-factor models. The experimental results show that the DeepForest-CQP-based multi-factor stock selection model achieves significant performance advantages in all backtesting metrics.

The empirical test of Asset Pricing Models in US Machinery market: Comparison of CAPM, FF3 and FF5 Models

As a sample, this study looks at data relating to the machinery sector of the US stock market from July 1926 to July 2022. This study's overarching objective is to investigate how three different financial models—the CAPM, the Fama-French 3 Factor Model, and the Fama-French 5 Factor Model, which have been implemented in the machinery sector over the course of time. This will allow the researchers to determine whether or not these models are applicable to the US stock market. In general, the three-factor model has better explanatory power for excess returns than the five-factor model, even with the addition of the investment and profitability factors for a single sector in the United States, the machinery sector and it explains almost all of the excess returns. This maybe because the three-factor model takes into account fewer variables than the multifactor model. Furthermore, the new profitability and investment factors do not have a single-factor effect that is as strong as that of the size and valuation factors.