Journal Article
Large Language Models

Comprehensive Review of Artificial Neural Network Applications to Pattern Recognition

Oludare Isaac Abiodun(Universiti Sains Malaysia), Muhammad Ubale Kiru(Lovely Professional University), Aman Jantan(Universiti Sains Malaysia), Abiodun Esther Omolara(Ahmadu Bello University), Kemi Victoria Dada(Ahmadu Bello University), Abubakar Malah Umar, OKAFOR UCHENWA LINUS(Nigerian Defence Academy), Humaira Arshad(Islamia University of Bahawalpur), Abdullahi Aminu Kazaure(The Federal Polytechnic, Ado-Ekiti), Usman M. Gana(Ibrahim Badamasi Babangida University)
January 1, 2019IEEE Access735 citations
PDFDOI

735

Citations

12

Influential Citations

IEEE Access

Venue

2019

Year

Abstract

The era of artificial neural network (ANN) began with a simplified application in many fields and remarkable success in pattern recognition (PR) even in manufacturing industries. Although significant progress achieved and surveyed in addressing ANN application to PR challenges, nevertheless, some problems are yet to be resolved like whimsical orientation (the unknown path that cannot be accurately calculated due to its directional position). Other problem includes; object classification, location, scaling, neurons behavior analysis in hidden layers, rule, and template matching. Also, the lack of extant literature on the issues associated with ANN application to PR seems to slow down research focus and progress in the field. Hence, there is a need for state-of-the-art in neural networks application to PR to urgently address the above-highlights problems for more successes. The study furnishes readers with a clearer understanding of the current, and new trend in ANN models that effectively addresses PR challenges to enable research focus and topics. Similarly, the comprehensive review reveals the diverse areas of the success of ANN models and their application to PR. In evaluating the performance of ANN models, some statistical indicators for measuring the performance of the ANN model in many studies were adopted. Such as the use of mean absolute percentage error (MAPE), mean absolute error (MAE), root mean squared error (RMSE), and variance of absolute percentage error (VAPE). The result shows that the current ANN models such as GAN, SAE, DBN, RBM, RNN, RBFN, PNN, CNN, SLP, MLP, MLNN, Reservoir computing, and Transformer models are performing excellently in their application to PR tasks. Therefore, the study recommends the research focus on current models and the development of new models concurrently for more successes in the field.

Analysis

Why This Paper Matters

This comprehensive review from 2019, with 735 citations, serves as a foundational reference for AI practitioners working on pattern recognition. It systematically catalogs the major ANN architectures—from classic MLPs to emerging Transformers—and evaluates their applicability to pattern recognition tasks. The paper is particularly valuable for its identification of persistent challenges, such as whimsical orientation (the inability to accurately calculate unknown directional paths) and the opaque behavior of neurons in hidden layers. By highlighting these gaps, the review provides a roadmap for future research, urging the community to both refine existing models and innovate new ones.

For practitioners, the paper offers a concise taxonomy of models and their typical use cases, making it easier to select appropriate architectures for specific pattern recognition problems. Its emphasis on statistical performance indicators (MAPE, MAE, RMSE, VAPE) also provides a common language for evaluating and comparing models, which is critical for reproducible research.

Technical Contributions

  • Comprehensive Model Taxonomy: The review categorizes 13 ANN architectures (GAN, SAE, DBN, RBM, RNN, RBFN, PNN, CNN, SLP, MLP, MLNN, Reservoir computing, Transformer) and discusses their strengths in pattern recognition.
  • Identification of Unresolved Problems: Highlights specific challenges including whimsical orientation, object classification, location, scaling, and template matching, which are often overlooked in application-focused papers.
  • Performance Metrics Framework: Standardizes evaluation using MAPE, MAE, RMSE, and VAPE, enabling cross-study comparisons.
  • Research Direction Recommendation: Advocates for concurrent development of current and new models, rather than exclusive focus on either.

Results

The paper does not present original experimental results but synthesizes findings from the literature. It reports that current ANN models—particularly GANs, CNNs, and Transformers—perform excellently in pattern recognition tasks. However, it does not provide concrete metrics or comparisons between models, as the review is qualitative. The key takeaway is that while many models achieve high performance, fundamental issues like whimsical orientation and neuron behavior analysis remain unsolved, limiting further progress.

Significance

This review has had substantial impact, evidenced by its 735 citations. It serves as a go-to reference for researchers entering the field of ANN-based pattern recognition, offering a clear overview of the landscape and highlighting where efforts are most needed. By explicitly naming unresolved problems, it helps prevent duplication of effort and encourages targeted research. For AI practitioners, the paper provides a useful checklist of models to consider and pitfalls to avoid, making it a practical resource for both academic and industrial applications.

artificial-neural-networkspattern-recognitiondeep-learningconvolutional-neural-networksgenerative-adversarial-networkstransformerssurvey
TransformersScaling LawsNeural Networks

Related Papers