AN ONLINE TRANSFER LEARNING FRAMEWORK WITH EXTREME LEARNING MACHINFOR AUTOMATED CREDIT SCORINGE
Abstract : Automated Credit Scoring (ACS) is the process of predicting user credit based on historical data. It involves analyzing and predicting the association between the data and particular credit values based on similar data. Recently, ACS has been handled as a machine learning problem, and numerous models were developed to address it. In this paper, we address ACS issues concerning credit scoring in a batch of machine learning problems, namely, feature irregularities due to empty features in many records, class imbalance due to non-uniform statistical distributions of the records between classes, and concept drift due to changing statistical characteristics concerning certain classes and features with time.
? Future work should incorporate feature selection to handle dynamic changes concerning relevant features and high dimensional data.
? In addition, the developed framework should be evaluated on other machine learning fields that share the same issues concerning the credit scoring problem.
? Future work is to extend the developed algorithm to include optimization of the random weights between the input-hidden layer algorithm and to incorporate dynamic feature selection.
? ACS has been handled as a machine learning problem, and numerous models were developed to address it.
? Automated credit scoring performance has been assessed using various approaches specified in the literature. Some studies used the binary classification problem
? The problem of credit scoring is discussed in the literature from various aspects: the first one is the handling of the missing features, the second one is the concept drift awareness and handling, the third one is approaches based on single classifier and the last one is the ensemble learning based approaches.
? A multiple time scale ensemble classifiers and a novel sample-based online class imbalance learning procedure are proposed to handle the potential concept drift and class imbalance in the client credit assessment data streams
? Lastly, a weighted formula was proposed that comprehensively considers the class imbalance ratio of the sample’s category and the prediction difficulty
? The optimization has been proposed based on multi-population of genetic with enhancement of crossover, mutation, and adding niche points and migrations
? In the work of the support vector machine SVM wa used with the incorporation of incorporates a group penalty function in the SVM formulation in order to penalize the variables simultaneously that belong to the same group, assuming that companies often acquire groups of related variables for a given cost rather than acquiring them individually.
? Note that the STL method only utilizes the credit data variables; that is, the training data label information is not used; therefore, the STL model is basically an unsupervised learning method.
|