hamideh razavi

The aim of this paper is to predict financing methods to support decision-making for startup founders and their investors. Initially, factors influencing the choice of financing methods, including structural, demographic, and behavioral factors, were identified. These factors were then assessed using a questionnaire consisting of 32 items, which was sent online to startup founders. Based on 70 responses received and using algorithms including binary matching, classification chains, label power set, K-nearest neighbors, extreme gradient boosting, cluster boosting algorithm and random forest, the financing methods chosen by startups were predicted. Comparison of the results from the algorithms shows that the boosting ensemble algorithm, with an F1 score of 89 and precison of 85%, predicts the selected financing methods on the test dataset better than other algorithms. Additionally, data analysis indicates that startups are more inclined towards personal funding methods, which aligns with the prevalence of loss aversion bias among entrepreneurs. Following loss aversion, overconfidence, anchoring, and illusion of control biases were the most frequent among entrepreneurs.

Overdue claims resulting from the lending process can pose a significant credit risk to financial institutions. To mitigate this risk, institutions often acquire guarantees. However, borrowers may encounter challenges when providing adequate and valid guarantees, particularly guarantees with lower risk. The present research focused on loan credit risk, borrower utility, and liquidity risk of guarantees within a private fund. First, data mining and classification methods were applied to a dataset of loans. The random forest algorithm, with a prediction accuracy of 0.986, was found to be optimal for constructing a guarantees composition model. The guarantees composition involves using multiple types of guarantees to secure a loan. Two models were established to generate guarantee compositions with a maximum default rate of 10%. In testing scenarios, the average risk of total default for acceptable combinations stands at 3.94%, a significant improvement compared to the fund loans’ overall default rate of 6.3%. Furthermore, the proposed model increases borrower utility from 4.22 to 4.6, not only reducing the default rate but also enhancing borrower utility.IntroductionWhen providing loans to customers, banks require guarantees due to insufficient knowledge of customers and the default risk. Obtaining guarantees from borrowers is recognized as a solution to reduce default risk in banks, but its impact on risk reduction depends on various factors. The combination and type of guarantees are among these factors, which have received less attention in the literature.The current state of overdue bank claims in Iran is unfavorable, and if conditions persist, it will lead to significant monetary and financial crises with negative effects on various sectors of the economy. In recent years, the ratio of non-performing loans to total disbursed facilities in Iran has been consistently higher, averaging around 5% to 10% higher than the global average. Reduction of the default risk in loans can decrease the ratio of non-performing loans to total disbursed facilities.The present study first intended to create various combinations of guarantees for each loan, followed by predicting the probability of default for each combination. In line with their priorities, borrowers can then select their desired guarantee composition from a list of acceptable combinations.MethodologyTo address the research problem, the study identified common classification methods in data mining by relying on published articles in the field of credit risk. Then, a sample dataset of loans from a financial institution was examined, and the data mining process based on classification methods was applied to the dataset. The random forest method, with a prediction accuracy of 0.986, was ultimately chosen as the approach for constructing the guarantee composition model. Using the previous guarantee compositions, the study developed two models by relying on machine learning techniques. These compositions take into account the perspectives of both the financial institution and the borrower.Final ResultThe two models generate guarantee compositions with a maximum acceptable default rate of 10%. Considering their own priorities circumstances, borrowers can select their desired guarantee composition from the available combinations, which contributes to a reduction in the default rate in the financial institution.

In time series clustering, features are typically extracted from the time series data and used for clustering instead of directly clustering the data. However, using the same set of features for all data sets may not be effective. To overcome this limitation, this study proposes a five-step algorithm that extracts a complete set of features for each data set, including both direct and indirect features. The algorithm then selects essential features for clustering using a genetic algorithm and internal clustering criteria. The final clustering is performed using a hierarchical clustering algorithm and the selected features. Results from applying the algorithm to 81 data sets indicate an average Rand index of 72.16%, with 38 of the 78 extracted features, on average, being selected for clustering. Statistical tests comparing this algorithm to four others in the literature confirm its effectiveness.

Parametric models are considered the widespread methods for time series forecasting. Non-parametric or machine learning methods have significantly replaced statistical methods in recent years. In this study, we develop a novel two-level clustering algorithm to forecast short-length time series datasets using a multi-step approach, including clustering, sliding window, and MLP neural network. In first-level clustering, the time series dataset in the training part is clustered. Then, we made a long time series by concatenating the existing time series in each cluster in the first level. After that, using a sliding window, every long-time series created in the previous step is restructured to equal-size sub-series and clustered in the second level. Applying an MLP network, a model has been fitted to final clusters. Finally, the test data distance is calculated with the center of the final cluster, selecting the nearest distance, and using the fitted model in that cluster, the final forecasting is done. Using the WAPE index, we compare the one-level clustering algorithm in the literature regarding the mean of answers and the best answer in a ten-time run. The results reveal that the algorithm could increase the WAPE index value in terms of the mean and the best solution by 8.78% and 5.24%, respectively. Also, comparing the standard deviation of different runs shows that the proposed algorithm could be further stabilized with a 3.24 decline in this index. This novel study proposed a two-level clustering for forecasting short-length time series datasets, improving the accuracy and stability of time series forecasting.

Despite contribution to catastrophic accidents, human errors have been generally ignored in the design of human-machine (HM) systems and the determination of the level of automation (LOA). This paper aims to develop a method to estimate the level of automation in the early stage of the design phase considering both human and machine performance.

A quantitative method is used to evaluate the performance of the whole human-machine system by the human-in-the-loop fault tree analysis while a qualitative and cross-sectional method is used to estimate human errors using the CREAM technique. The data are collected from real cases that happened in the control room of the Ferdowsi power plant.

Full automatic option with an average error of 0.013 had the lowest error rate, i.e. 1/8 of the error rate of the manual design. In addition, the CREAM analysis showed that the control room operators were not satisfied with the availability of procedures and Man-Machine Interface and operational support in general. Thus, on average, the reliability of the manual design is less than the reliability of the automatic setting.

High machine reliability has led to the fact that the fully automatic design would be one of the best design choices for human-machine systems. However, based on the previous studies, high automation may have some human-out-of-the-loop shortcomings. Thus, this study proposed solutions to overcome these disadvantages based on the importance of the control parameters or the essence of human involvement in some decision-making and execution tasks.

In some practical applications, the quality of a process or a product is characterized by a relationship, namely a profile between a response variable and one or more explanatory variables. Assuring the capability of a profile to meet the process requirements is particularly important. Process capability indices (PCIs) are widely used to measure whether a process is capable of reproducing product items within the specification limits (SLs). Background literature on the PCIs for profiles mostly take crisp values for process data. However, in practice, the outcomes of a measurement are often imprecise. So, the basic assumption of crisp data for process capability analysis (PCA) in profiles is not valid. Hence, fuzzy methods are developed to analyze the capability of a fuzzy profile with fuzzy response data. To this end, we extend the functional approach based on fuzzy set theory for the situations in which the SLs and target values of the response variable are imprecise. The performance of the proposed indices is investigated through simulation studies. The simulation results confirm that the proposed method performs well regarding -distance between the estimated value and the true value of fuzzy PCIs. Furthermore, a case study shows the applicability of the proposed method.

In industrial and manufacturing processes, there are several input factors in different levels, which may affect the final product characteristics. The design of experimental method, as one of the new quality improvement methods identifies the most important effective factors factors on product quality by experimental design and adjusts its optimal levels. The purpose of this study was determination of the optimum weight of the 500-ml serum bottle using the 2k design of experimental, so that, if thicker, it would increase the consumption of the primary substance of P.P. And losses, and if it is used less than standard, it will cause deformation of the body and loss of the product. Because the serum production process faces some inevitable waste but subsequent changes in weight cause the consumption control not be properly performed, so after determining the weight the standard of the serum the conditions of the machine must be checked, it is possible to approach the production with a standard weight and the normal waste is under control. In this study, we tried to identify the factors affecting the weight of the serum bottle using 2k experimental design. After analyzing and evaluating the adequacy of the model, the results of analysis of variance and main effects graphs showed that the timing of cycle, the temperature of the heaters and the extruder circle had the most direct effect on the weight of the bottle. By controlling these factors, we can determine the weight fluctuations reduced the bottle.

Facility Layout Problem (FLP) is an important issue in plant design, which directly affects production costs as well as safety concerns. Hierarchical and integrated approaches are two main methods used for the formulation of FLP. In the hierarchical approach, FLP is solved after the completion of the process design and scheduling, while the integrated approach treats FLP simultaneously with other phases. This paper presents a new approach based on the candidate groups for interactive functions; incorporating the advantages of the previous approaches.
In the new approach, the design phases prior to FLP results in a set of models introducing by the engineering team in advance. Next, the models are solved such that instead of individual designs, a set of suggested structures are generated. These structures are then fed into the FLP as input data and solved by mixed integer programming technique. Some characteristics of the models are: presence of the hub stations, different and variable dimensions for each station, possibility of rotation for the stations and incorporating safety distances. Finally, test problems are solved and results are discussed. It should be noted that if a candidate group entails a different production flow then a different optimum solution might be obtained.

Facility layout problems have been generally solved either hierarchically or integrated into other phases of plant design. In this paper, a hybrid method is introduced so that clustering and facilities layout can be simultaneously optimized. Each cluster is formed by a group of connected facilities and selection of the most appropriate cluster configuration is aimed. Since exact method by MIP is limited to small problems, a heuristic algorithm including constructive and improving phases is developed. In order to enhance the performance of the algorithm, systematic generation of intersection points inside available area together with shaking, split groups and Tabu list techniques are used.Then, two different examples are presented and the comparison of the results supports the merit of the proposed algorithm. For further validation, 18 test problems are solved both by the proposed algorithm and MIP by CPLEX. Comparison of the results reveals that for up to 13 facilities, the best solutions of the algorithm are equal to optimum solution of MIP but achieved in shorter times. For larger problems with higher number of facilities, even though processing times for MIP is much longer, in almost all cases, it cannot produce the best solutions of the proposed algorithm.

9% of all available packaging are jars with steel lids. However some consumers have troubles with jar opening. The aim of the current research is the analysis of the exerting force onto the jar lids relating the torque input to the anthropometric and physiological characteristics of the consumers.
Material and In this study, we examine the force distributed on the fingers and the palm using a special tester consists of the force sensing resistors (FSR) sensors. This tester includes 19 sensors, 14 of which are installed on the fingers and 5 on the palm. Each sensor is connected to a driver via two wires and then to a computer by a USB port. A full jar with 1100g weight and 85mm lid outer diameter was used and the forces exerted by 20 participants including 10 male and 10 female students were recorded. Then Minitab 17 software was used and the mathematical relationship between the torque and anthropometric and physiological characteristics was investigated.
It was found that the force is mainly exerted by the index finger. The calculation of the torque based on the force measurements showed that the maximum torque applied by men was 10.63Nm and by women was 7.66Nm. Statistical F-test for the torque (response variable) at 0.05 significance level resulted in p-values equal to 0.011 and 0.008 for male and female participants, respectively. Therefore, it can be concluded that the measured values are statistically significant.
Statistical analysis revealed that the hand dimensions and the grip strength play the main roles in applying torque. Additionally, the results indicated that packaging conditions for the particular jar in this survey must be adjusted such that the consumers with the lowest physical ability (5.11 Nm torque) can open them.

This study was designed to investigate the effect of mental fatigue on response-time and error-rate of cyclic tasks using a simulated model in ACT-R cognitive architecture. The effects of fatigue on cyclic tasks with high frequency and long duration have not thus far been systematically studied. In the present study، an unvarying cyclic task was modeled in ACT-R (Adaptive Control of Thought-Rational) environment. For each run، the average response-time per cycle and total error-rate were obtained through dividing the total working time and total number of incorrect responses، respectively by the total number of cycles after eliminating the learning phase. Then، the effect of mental fatigue was studied in the cognitive architecture and a fatigue-based multiplier was incorporated into the model. This multiplier was a concentration index which was declined when response-time and error-rate increased. Next، real data were collected through experiments on volunteers and compared with simulated results using analysis of variance (ANOVA). After validation of the model، the impact of fatigue-based multipliers on the response-time and error-rate were analyzed for different iteration numbers. Using an F test at α=0. 05، no significant disagreement between virtual and real data and the validity of the model was shown. Final results for 3-sec cycle tests revealed up to 0. 5 sec increase in the average response-time for each cycle and 12% in the average error-rate. After 1000 iterations of the sample task، the average response-time for each cycle and the average error-rate are increased by 0. 43 sec and 8%، respectively. This confirms the negative impact of mental fatigue on such indices. Such an effect can be administered and controlled by proper rest breaks.

Ethosomes are a novel and alternative drug delivery systems and currently the focus of many research activities. Ethosomes are soft, flexible vesicles that are composed mainly of phospholipids, ethanol and water. Ethosomes can provide better skin permeation than liposomes and deliver enhanced amounts of both small and large therapeutic agents through skin and targeting to deeper skin layers for various diseases. Finally deliver to the systemic circulation. It is clear that ethosomal carrier because of presences ethanol cause skin disruption and increases lipid fluidity. In this review, the mechanism of penetration, applications, preparation, advantages and disadvantages of ethosomes are illuminated.

A sensitive monitoring of biological analytes, such as biomolecules (protein, lipid, DNA and RNA), and biological cells (blood cell, virus and bacteria), is essential to assess and avoid risks for human health. Nanobiosensors, analytical devices that combine a biologically sensitive element with a nanostructured transducer, are being widely used for molecular detection of biomarkers associated with diagnosis of disease and detection of infectious organisms. Nanobiosensors show certain advantages over laboratory and many field methods due to their inherent specificity, simplicity and quick response. In this review, recent progress in the development of nanobiosensors in medicine is illuminated. In addition, this article reviews different kinds of bio-receptors and transducers employed in nanobiosensors. In the last section, overview of the development and application of various nanomaterials and nanostructures in biosensing has been provided. Considering all of these aspects, it can be stated that nanobiosensors offer the possibility of diagnostic tools with increased sensitivity, specificity, and reliability for medical applications.

Irrational use of the drug has high impression on the household economy, treatment and health. A lot of attention must be paid to the Medical group which is the most effective elements in the process of prescribing medication. Although awareness of doctors in the principles of prescribing is in acceptable range, different results can be seen in the implementation. Therefore it is necessary to compare the effective methods of changing the prescribing behavior. In this training study, 50 general practitioner (GP) that have offices and most of the indicators of irrational drug prescription were selected for an intervention. After the initial assessment and matching, based on the aim related sampling, they were divided into two groups affected by health belief model different training structures. Finally after a year their performance were collected and compared with application processing and statistical software version. According to the nondependent t-test the number of drugs per prescription between two groups did not show significant difference. However, administration of antibiotics and injectable form of drugs in the training group of perceived benefits and barriers was reduced as (p≤0.05) compared to training group of susceptibility and perceived severity. Also drug interactions into two groups have not significant difference. Training the perceived benefits and barriers are probably more effective than training the susceptibility and perceived severity in changing the function and behavior of the target group.

In this paper, a new approach to team staffing is proposed using effectiveness criteria and PROMETHEE method. Initially, the most important criteria are identified through background studies, the questionnaires and interviews, and their comparison. These criteria are both quantitative and qualitative and can be classified in four groups including: personal, team working, organizational and occupational; and can be modified based on the nature and the virtue of the company. In addition to the criteria, scaled weights and the candidates’ specifications constitute the input to the model. Next, PROMETHEE technique in Decision Lab software is applied for solving the model. Since the weights of the criteria directly affect on the responses, the technique works straightforward and efficient. It promptly specifies an order of the best choices of members for the team which is perfectly in agreement with the arrangement suggested by the head of the team. This model is finally verified with GAHP method and the validation of the results is confirmed.

In many applications the quality of a process or product is best characterized and summarized by a functional relationship between a response variable and one or more explanatory variables. Profile monitoring is used to understand and to check the stability of this relationship over time. Therefore checking the capability of a profile seems essential. Process capability indices have become very popular in assessing the capability of manufacturing processes in practice during the past decade. In this paper, a method is proposed for determining the capability of a linear profile by means of calculating the capability indices for intercept and slope, independently. The intercept measures are explanatory variables that can be treated the same as single quantities. Therefore, the problem is reduced to evaluating the process capability for slope of the profile. Since the overall slope is a linear regression on a series of subsequent observations, the mutual gradients provide a set of variables for process capability study. After checking the normality and control status of the slope variables, PCRs can be evaluated for sample data