Relationship between dependent binary variables and ration level independent variables.

Answer

Question #1(a)

The best method for analysis of the case presented is the unsupervised analytical method. The technique requires definite response measures to enhance predictive analytics. The organization has past data on the loan application and repayment history of clients and whether they were approved (Huang, Wang, Zheng, Tang, Liu, & Hong, 2020). The data will be used to understand the behavior of loan applicants to determine if they qualify for a loan or not. The method uses clustering that involves groups of data points with natural similarities. In this case, the clustering of loan applicants will reveal data that was initially hidden. Thus, it would enable data mining operations that can lead to the discovery of new correlations.

 Nonetheless, the technique would be suitable for use because of its ability to conduct association. The system identifies common co-occurrences in a group of events. For instance, I will categorize loan applicants into various categories such as defaulters, those whose loans were declined, and those whose loans were approved to provide an accurate report to loan officers. Information from the association could be used to personalize customers' experience and promote services that increase the interest of customers with the ability to apply for loans. Besides, it will help the loan officers track customers' repayment performance and prevent them from more suitable loan options (Huang, Wang, Zheng, Tang, Liu, & Hong, 2020). The featured extraction attribute of the technique creates new features based on the attributes of the data. Categorizing data points by their attributes helps in compressing the data. Thus, it makes it possible to make predictions and recognize the patterns. The technique will assist in predicting the behavioral pattern of loan applicants based on the information on the history of their applications.

Question #1(b)

Logistic regression would be appropriate in the case presented. The technique offers a predictive analysis by using the existing data to make meaningful inferences. It is used in data description to explain the relationship between dependent binary variables and one or more nominal, interval, ordinal, and ration level independent variables. As a financial analyst, I want to provide accurate predictive information to the loan officers to help them understand the behavior of the customers (Mansournia, Geroldinger, Greenland, Heinze, 2018). Thus, I would use regression since it does not require an assumption of normality. Besides, it is suitable since it can handle discrete or independent continuous variables. The computation of customer data will include various types of figures, both continuous and discrete data. Primarily, logistic regression would be used in this case since it helps in the analysis of the effects of explanatory factors in relative risks of the outcome. Thus, logistic transformation gives information as logarithms that indicate success or failure. Nonetheless, the analysis would involve the classification of customers based on their loan borrowing abilities and this can only be achieved effectively by logistic regression. Linear regression would not be suitable for analysis in this case since it only predicts an absolute number ranging from0-1(Kim, Song, Wang, Xia, & Jiang, 2018). The case requires an analysis that explains the probability of an event occurring, with a probability range from 0-1. Thus, I would choose to apply logical regression over linear regression in this case.

Question #1 (c)

Dependent Variable

1. Defaulting
2. Repayment

• Declined Loan application

Defaulting is dependent on the ability of the applicant to pay his or her loan successfully. Loan awarded due to repayment history depends on the previous repayment. That is, the loan officers can either approve or fail to approve the loan depending on the repayment status of the customer. Declined loan applications will depend on the ability of the applicant to repay the loan or the previous loan repayments history.

Independent Variable

1. Loan

The loan is the independent variable in this case, and it is used to predict the dependent variable. It stands on its own and is not affected by the dependent variable in any way.

Question #2

The type of data management technology that would be most applicable for the company is a relational database. The database organizes data into tables in a manner that they can be linked based on the similarities of the data. The company keeps the information about customers, types of food, and purchase history (Setyawati, Wijoyo & Soeharmoko, 2020). Thus linking the information can provide important data to analyze the characteristics of the customers of the organization. Primarily, the relationship database allows will allow the organization to understand the relationship among available data and gain insights for making relevant decisions and identifying new opportunities to expand the operations of the firm.  For instance, the organization can develop a customer table that contains the customers’ food preferences and a transaction table with data describing individual transactions. The technique, therefore, creates useful information by joining the tables resulting in an understanding of the relationship between the data. The analysis can order results using data name or any column. The database is flexible, reduces redundancy, and provides eases of backup and disaster recovery.