Object and goal

• This is our first work utilizing a transformation method that converts numerical data into images. In this research, we focused on exploring the feasibility and forecasting efficiency of NISE using YOLO. The results are presented in our published paper. Performance speed was not a primary consideration in this initial version, as our main goal was to demonstrate the concept. The processing speed depends largely on the YOLO model used, as noted in our discussion. We are currently working on improvements for future versions to enhance efficiency.
• In the specific case of a more efficient and faster prediction implementation that we have not reported in this paper, after you run strip_pre.py you can go to our implementation on Google, which uses fewer epochs and is faster only for all types of numeric forecasting. In the case of text or other features that are not the first one, we will develop it in the future.

Download Links

• Paper Available:
  Read our published paper on Springer
  If you find this work helpful, please consider citing our paper.
• Download Program (yolov5_strip.zip) We note that our project is run on windows but we think a little modification, it can be run on Linux too.
  How to Run:
  Start by running strip_pre.py. The code contains step-by-step comments (steps 0 to 3) explaining how to create the NISE.
  Note: The requirements file (python setup package > requirements.txt) is not yet prepared. However, all libraries used are standard and can be installed via pip.
  We apologize for any inconvenience.
  Required Libraries (We use 2 files to generate the NISE image: 1.strip_pre.py and 2.lib.py):

  • At file: strip_pre.py
  • os
  • json
  • At file: lib.py
  • statistics
  • cv2 (pip install opencv-python)
  • numpy
  • PIL (Pillow)
  • sklearn (decomposition, preprocessing, metrics)
  • shutil
  • pathlib
  • glob
  • collections.Counter
  • math
  • pandas
  • matplotlib

  strip_pre.py will generate 5 k-fold image sets in the folders for 3 algorithms: output_k_fold, output_k_fold2, output_k_fold3 which produce ready to use for training, validating and testing for YOLOv5 format.
• Download Weight
  The training weight and image k-folds for training of every dataset and the 3 algorithms following the paper are here.
• Download Data in csv (data.zip)
  Dataset is from UCI (wine, iris, student performance and breast cancer) as we cited in our paper.
• Future Work
  • We plan to use MobileNet as our base model for improved efficiency and performance.
  • We are currently testing NISE on Google Teachable Machine using the MobileNet architecture which we think in our case it be better performace. It is also look possible and also can be useable in actual work.
  
• References
  1. Africa, A.D., Velasco, J.: Development of a urine strip analyzer using artificial neural network using an android phone. ARPN Journal of Engineering and Applied Sciences 12, 1706–1713 (03 2017)
  2. Bochkovskiy, A., Wang, C.Y., Liao, H.Y.M.: Yolov4: Optimal speed and accuracy of object detection. ArXiv abs/2004.10934 (2020), https://api.semanticscholar.org/CorpusID:216080778
  3. Cortes, C., Vapnik, V.: Support-vector networks. Machine learning 20(3), 273–297 (1995)
  4. Cortez, Paulo Cerdeira A. Almeida F., M.T., J., R.: Wine Quality. UCI Machine Learning Repository (2009)
  5. Cover, T., Hart, P.: Nearest neighbor pattern classification. IEEE Transactions on Information Theory 13(1), 21–27 (1967)
  6. El Ghmary, M., Ouassine, Y., Ouacha, A.: License Plate Character Recognition System Using YOLOv5, pp. 588–594 (03 2023)
  7. Fisher, R.A.: Iris. UCI Machine Learning Repository (1936)
  8. Jocher, G.: Ultralytics yolov5 (2020), https://github.com/ultralytics/yolov5
  9. Kim, S.C., Cho, Y.S.: Predictive system implementation to improve the accuracy of urine self-diagnosis with smartphones: Application of a confusion matrix-based learning model through rgb semiquantitative analysis. Sensors 22, 5445 (07 2022)
  10. Mishra, S., Sarkar, U., Taraphder, S., Datta, S., Swain, D., Saikhom, R., Panda, S., Laishram, M.: Principal component analysis. International Journal of Livestock Research p. 1 (01 2017)
  11. Ng, W.K., Choi, S., Ravishankar, C.: Lossless and lossy data compression (06 1998)
  12. Pei, C.: Research on the influencing factors of student performance. Theoretical and Natural Science 51, 26–33 (09 2024). https://doi.org/10.54254/2753-8818/51/2024CH0131
  13. Ragab, M., Jadid Abdulkadir, S., Muneer, A., Alqushaibi, A., Sumiea, E., Qureshi, R., Al-Selwi, S., Alhussian, H.: A comprehensive systematic review of yolo for medical object detection (2018 to 2023). IEEE Access PP, 1–1 (01 2024)
  14. Rajpurkar, P., Hannun, A.Y., Haghpanahi, M., Bourn, C., Ng, A.Y.: Cardiologist-level arrhythmia detection with convolutional neural networks (2017), https://arxiv.org/abs/1707.01836
  15. Raza, A., Shaikh, M., Siddiqui, O., Ali, A., Khan, A.: Enhancing agricultural pest management with yolo v5: A detection and classification approach. UMT Artificial Intelligence Review 3 (12 2023)
  16. Redmon, J., Divvala, S., Girshick, R., Farhadi, A.: You only look once: Unified, real-time object detection. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 779–788 (2016)
  17. Redmon, J., Farhadi, A.: Yolov3: An incremental improvement. ArXiv abs/1804.02767 (2018), https://api.semanticscholar.org/CorpusID:4714433
  18. Wolberg, William, M.O.S.N., Street, W.: Breast Cancer Wisconsin (Diagnostic). UCI Machine Learning Repository (1993), DOI: https://doi.org/10.24432/C5DW2B
  19. Wu, X., Kumar, V., Quinlan, J.R., Ghosh, J., Yang, Q., Motoda, H., McLachlan, G.J., Ng, A., Liu, B., Philip, S.Y., et al.: Top 10 algorithms in data mining. Knowledge and information systems 14(1), 1–37 (2008)
  20. Zhang, Y., Guo, Z., Wu, J., Tian, Y., Tang, H., Guo, X.: Real-time vehicle detection based on improved yolo v5. Sustainability 14, 12274 (09 2022)