Assessing the predictive performance of artifIcial neural network-based classifiers based on different data preprocessing methods, distributions and training mechanisms

Costea, Adrian; Nastac, Dumitru Iulian

doi:10.1002/isaf.269

Cited by 5 publications

(4 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They also investigated the implication of three types of crossover operator, such as uniform, arihtmetic and one-point crossover, on the prediction performance of GA-based ANN and found no significant difference between these different crossover operators. In Costea and Nastac (2005), the authors used GA-based ANN training to improve the performance of an already trained ANN classification model, but the genetic algorithm was unable to improve the accuracy of the ANN model. This result might be due to the high accuracy rate of the trained network that the genetic algorithm had to improve (in excess of 90%).…”

Section: Second Stage Of the Methodology: Building The Performance CLmentioning

confidence: 99%

“…The generic classification model based on a neural approach, adapted from Costea and Nastac (2005) is based on some preliminary steps: pre-processing data and separating data into training (TR) and test (TS) sets. After that, the proper ANN architecture is constructed, by determining the proper number of hidden layers and the appropriate number of neurons in each hidden layer.…”

Section: Second Stage Of the Methodology: Building The Performance CLmentioning

confidence: 99%

See 1 more Smart Citation

An Integrated Two-Stage Methodology for Optimising the Accuracy of Performance Classification Models

Costea

Феррара

Şerban

2017

Technological and Economic Development of Economy

Self Cite

View full text Add to dashboard Cite

In this paper we propose a two-stage methodology to classify the non-banking financial institutions (NFIs) based on their financial performance. The first stage of the methodology consists of grouping the companies in similar financial performance classes (e.g.: “good”, “average”, “poor” performance classes). We optimise the allocation of the observations within the performance clusters by applying an enhanced version of an observation re-allocation procedure proposed in our previous work. Next, based on the result of the grouping phase, we construct a performance class variable by attaching a performance label to each data row. Then, in the second phase of our methodology, we propose a feed-forward neural-network classification model that maps the input space to the newly-constructed performance class variable. This model allows us to forecast the performance of new companies as data become available.

show abstract

Section: Second Stage Of the Methodology: Building The Performance CLmentioning

confidence: 99%

Section: Second Stage Of the Methodology: Building The Performance CLmentioning

confidence: 99%

An Integrated Two-Stage Methodology for Optimising the Accuracy of Performance Classification Models

Costea

Феррара

Şerban

2017

Technological and Economic Development of Economy

Self Cite

View full text Add to dashboard Cite

show abstract

“…Each eigenvalue needs to be transformed in the order of magnitude to make it within the range of (0, 10). This range is selected based on the existing research on data pre-processing [18,19].…”

Section: An Improved Methods Of Data Pre-processing For Classificationmentioning

confidence: 99%

Big Data Mining Optimization Algorithm Based on Machine Learning Model

Jiao¹

2020

RIA

View full text Add to dashboard Cite

In a big data mining optimization algorithm, the classification algorithm plays an important role. At present, there are many popular classification algorithms based on machine learning. Aiming at the problems of existing big data classification algorithms, two improved strategies and implementation methods are proposed in this paper. First of all, before the data training, the orders of magnitude of the original data are normalized to achieve better data pre-processing and classification. Then with the least mean square algorithm and the BP neural network classification algorithm as the basic algorithms, an improved batch learning BP algorithm is designed, based on the rules of batch learning. The experimental results indicate that the improved batch learning BP algorithm can better solve the imbalanced classification problem in big data.

show abstract

“…Thus, MLP is the most popular neural network method that has been widely used for many practical applications, and one good reason is that able to learn non-linear representations. It has been widely employed for modeling, prediction, classification, clustering, and optimization purposes (Ahmed, 2005;Bose, 2007;Costea & Nastac, 2005;De Gooijer & Hyndman, 2006;Do, Taherifar, & Vu, 2019;Ramchoun, Idrissi, Ghanou, & Ettaouil, 2017;Zacharis, 2016).…”

Section: Market Segmentation and Neural Networkmentioning

confidence: 99%

Recognizing Saltwater Recreational Angers’ Motivations Using Multilayer Perceptron Neural Network

Nain

2022

ijsar

View full text Add to dashboard Cite

The purpose of this study was to examine saltwater recreational anglers’ answers to the fifteen statements regarding the importance of fishing trips, and to classify groups exhibiting common patterns of responses from individuals’ recreational fishing motivations using the data extracted from the database collected from the 2013 National Saltwater Angler Survey. Using the factor analysis, the fifteen statements were reduced into five dimensions, named catch, information, site preferences, social, and management. Empirical results based on the k-means clustering analysis identified three different saltwater recreational angler groups, named catch and social, site choice, and fishing related groups. Results of the discriminant analysis indicated that cluster means were significantly different. The multilayer perceptron neural network model was utilized as a predictive model in deciding the classification of saltwater anglers based on recreational fishing motivations. From an architectural perspective, it showed a 15-9-3 neural network construction. This study may provide insight into the information about what types of saltwater recreational anger groups exist and identifying unknown groups in the data set for saltwater recreational fishing planning and management purposes.

show abstract

Assessing the predictive performance of artifIcial neural network-based classifiers based on different data preprocessing methods, distributions and training mechanisms

Cited by 5 publications

References 42 publications

An Integrated Two-Stage Methodology for Optimising the Accuracy of Performance Classification Models

An Integrated Two-Stage Methodology for Optimising the Accuracy of Performance Classification Models

Big Data Mining Optimization Algorithm Based on Machine Learning Model

Recognizing Saltwater Recreational Angers’ Motivations Using Multilayer Perceptron Neural Network

Contact Info

Product

Resources

About