IMPLEMENTING MACHINE LEARNING IN DATA CLASSIFICATION

Notice

This is an unedited manuscript accepted for publication and provided as an Article in Press for early access at the author’s request. The article will undergo copyediting, typesetting, and galley proof review before final publication. Please be aware that errors may be identified during production that could affect the content. All legal disclaimers of the journal apply.

Year : 2025 | Volume : 03 | 02 | Page :
    By

    Monalisa Hati,

  • Khurram Rashid,

  1. Assistant Professor, Department of Computer Science and Engineering, AMITY School of Engineering and Technology, AMITY University, Mumbai, Maharashtra, India
  2. Research Scholar, Department of Computer Science and Engineering, AMITY School of Engineering and Technology, AMITY University, Mumbai, Maharashtra, India

Abstract

Data classification forms an essential aspect of artificial intelligence (AI) and soft computing, helping a great deal in the transformation of raw data into knowledge that forms the basis of numerous applications, such as fraud detection, medical diagnostics, and natural language processing. This paper discusses the challenges and the state of the art in data classification, as far as scalability, noise handling, and feature selection optimization are concerned. It gives a review of classical classification methods such as decision trees, the support vector machine (SVM), and a host of ensemble learning algorithms vis-a-vis modern deep learning architectures like convolutional neural networks (CNN) and recurrent neural networks (RNN). Consequently, soft computing methods, such as fuzzy logic and genetic algorithms, are reviewed to ascertain how they can enhance performance concerning noisy, incomplete, or high-dimensional data. This paper describes how AI and soft computing can merge given hybrid models that combine neural networks with fuzzy systems hierarchy to improve classification accuracy and interpretability. The methodology begins with a description of, in particular, the most popular frameworks utilized for model development: TensorFlow, PyTorch, and MATLAB, along with hyperparameter tuning strategies such as grid search, random search, and Bayesian optimization. Evaluation metrics such as accuracy, precision, recall, F1 score, or AUC-ROC find their application in various use cases such as facial recognition or medical imaging and financial fraud detection to showcase the effect of the proposed techniques. From the results, hybrid methods performed better than the conventional model against noisy and complex datasets and actually impart extensiveness and adaptability to the models. The results of the case studies support improvements in terms of classification accuracy and robustness. In conclusion, future work involves automation of feature selection, exploring additional hybrid approaches, and addressing ethical issues such as fairness and transparency in classification systems.

Keywords: TensorFlow, F1 score, SVM, CNN, RNN

How to cite this article:
Monalisa Hati, Khurram Rashid. IMPLEMENTING MACHINE LEARNING IN DATA CLASSIFICATION. International Journal of Data Structure Studies. 2025; 03(02):-.
How to cite this URL:
Monalisa Hati, Khurram Rashid. IMPLEMENTING MACHINE LEARNING IN DATA CLASSIFICATION. International Journal of Data Structure Studies. 2025; 03(02):-. Available from: https://journals.stmjournals.com/ijdss/article=2025/view=210737


References

  1. Bishop, C. M. (2006). Pattern Recognition and Machine Learning. Springer. – Comprehensive guide on classification methods, including decision trees and SVMs.
  2. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press. – Detailed exploration of deep learning architectures such as CNNs and RNNs.
  3. Zadeh, L. A. (1965). “Fuzzy Sets.” Information and Control – Foundational paper on fuzzy logic, key to soft computing approaches.
  4. Hastie, T., Tibshirani, R., & Friedman, J. (2009). The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Springer. – Explains machine learning methods, including ensemble learning.
  5. LeCun, Y., Bengio, Y., & Hinton, G. (2015). “Deep Learning”. – Overview of deep learning applications in data classification.
  6. Labeled Faces in the Wild (LFW) – Dataset for facial recognition tasks.
  7. Credit Card Fraud Detection Dataset. Available at Kaggle – Public dataset for studying fraud detection.
  8. chest-xray-pneumonia – https://www.kaggle.com) – Dataset for medical imaging and classification tasks.
  9. Scikit-learn Documentation – Tool for machine learning model implementation.
  10. TensorFlow and Keras Documentation: https://www.tensorflow.org/- Frameworks for deep learning model development.
  11. MATLAB Fuzzy Logic Toolbox – Tools for implementing neuro-fuzzy systems.
  12. Titanic Dataset- https://www.kaggle.com/c/titanic) – Classic dataset for binary classification tasks.
  13. Iris Dataset – Dataset for multi-class classification problems.
  14. Colab Notebook on Classification Tasks. Accessible at https://colab.research.google.com- Codebase for experimentation and analysis.
Ahead of Print Subscription Review Article
Volume 03
02
Received 28/03/2025
Accepted 08/04/2025
Published 21/05/2025
Publication Time 54 Days
438

Login


My IP

PlumX Metrics