Düşme algılama sistemlerinin geliştirilmesinde PySpark ve Scikit-Learn kütüphanelerinin performansının araştırılması

Year 2024, Volume: 13 Issue: 2, 582 - 592, 15.04.2024

Erhan Kavuncuoglu

https://doi.org/10.28948/ngumuh.1388789

Abstract

Düşmeler, genellikle ciddi yaralanmalara ve yaşlı nüfusun yaşam kalitesinin azalmasına neden olan önemli bir risk oluşturur. Doğru ve etkili düşme tespit sistemleri, bu riskleri azaltmada önemli bir rol oynayabilir. Bu çalışma, düşme tespit modellerinin geliştirilmesinde PySpark ve Scikit-Learn kütüphanelerinin performansını karşılaştırmalı bir analiz sunmaktadır. Her iki kütüphane de kullanılarak, lojistik regresyon, gradyan arttırma sınıflandırıcısı, rastgele orman, destek vektör makinesi ve karar ağacı dahil olmak üzere beş popüler makine öğrenme algoritması kullanılarak düşme tespit modelleri oluşturuldu. Modeller, kapsamlı metrikler (doğruluk, duyarlılık, özgüllük, karışıklık matrisi) kullanılarak değerlendirildi. Çalışmada düşme ve günlük yaşam aktivite verilerinden oluşan Sisfall veri setinden 26 farklı özellik beş ana kategoride çıkarıldı: temel istatistiksel özellikler, frekans alanı özellikleri, zaman serisi özellikleri, hareket özellikleri ve ilişkisel özellikler. Bu özellikler, düşme tespit modellerine düşmeleri tanıma yeteneklerini artırmak için dahil edildi. Bulgular, hem PySpark hem de Scikit-Learn'ün düşme tespitinde güçlü ve etkili sonuçlar sunduğunu göstermektedir. Her iki kütüphane de en yüksek performans oranlarına lojistik regresyon ile ulaşılmıştır. Ayrıca, PySpark, testte daha iyi performans sergileyen Scikit-Learn'e göre biraz daha uzun eğitim süreleri sergilemiştir. Sonuç olarak, bu çalışma, yaşlıların güvenliğini ve refahını artırmak için düşme tespit sistemlerinin geliştirilmesine katkıda bulunduğu gibi yeni bir özellik çıkarma yöntemi sunarakta literatüre katkıda bulunuyor.

Keywords

Düşme Algılama, Yapay Zeka, Makine Öğrenmesi, PySpark, Scikit-Learn

Exploring the performance of PySpark and Scikit-Learn libraries in developing fall detection systems

Abstract

Falls pose a significant risk, often resulting in serious injuries and reduced quality of life for the elderly population. Accurate and effective fall detection systems can play an important role in reducing these risks. This study presents a comparative analysis of the performance of PySpark and Scikit-Learn libraries in the development of fall detection models. Using both libraries, fall detection models were built using five popular machine learning algorithms, including logistic regression, gradient boosting classifier, random forest, support vector machine and decision tree. The models were evaluated using comprehensive metrics (accuracy, sensitivity, specificity, confusion matrix). In the study, 26 different features were extracted from the Sisfall dataset consisting of falls and activities of daily living data in five main categories: basic statistical features, frequency domain features, time series features, motion features and relational features. These features were incorporated into the fall detection models to increase their ability to recognise falls. The findings show that both PySpark and Scikit-Learn offer powerful and effective results in fall detection. The highest performance rates of both libraries were achieved by logistic regression. Furthermore, PySpark exhibited slightly longer training times than Scikit-Learn, which performed better in the test. In conclusion, this study contributes to the development of fall detection systems to improve the safety and well-being of the elderly and contributes to the literature by providing a new feature extraction method.

Keywords

Fall Detection, Artificial Intelligence, Machine Learning, PySpark, Scikit-Learn

References

• A. Sucerquia, J. D. López, and J. F. Vargas-Bonilla, SisFall: A fall and movement dataset, Sensors, vol. 17, no. 1, Art. no. 1, Jan. 2017, doi: 10.3390/s17010198.
• M. Islam et al., Deep learning based systems developed for fall detection: A Review, IEEE Access, vol. 8, pp. 166117–166137, 2020, doi: 10.1109/ACCESS.2020.3021943.
• T. C. Nokeri, Principal component analysis with Scikit-Learn, PySpark, and h2o, in data science solutions with python: fast and scalable models using keras, PySpark, mllib, h2o, xgboost, and Scikit-Learn, T. C. Nokeri, Ed., Berkeley, CA: Apress, 2022, pp. 101–110. doi: 10.1007/978-1-4842-7762-1_9.
• T. C. Nokeri, Cluster analysis with Scikit-Learn, PySpark, and h2o, in data science solutions with python: fast and scalable models using keras, PySpark mllib, h2o, xgboost, and Scikit-Learn, T. C. Nokeri, Ed., Berkeley, CA: Apress, 2022, pp. 89–99. doi: 10.1007/978-1-4842-7762-1_8.
• M. Junaid et al., Performance evaluation of data-driven intelligent algorithms for big data ecosystem, Wirel. Pers. Commun., vol. 126, no. 3, pp. 2403–2423, Oct. 2022, doi: 10.1007/s11277-021-09362-7.
• T. C. Nokeri, Tree modeling and gradient boosting with Scikit-Learn, xgboost, PySpark, and h2o, in data science solutions with python: fast and scalable models using keras, PySpark mllib, h2o, xgboost, and Scikit-Learn, T. C. Nokeri, Ed., Berkeley, CA: Apress, 2022, pp. 59–74. doi: 10.1007/978-1-4842-7762-1_6.
• T. S. Patel, D. P. Patel, and C. N. Patel, Real time scalable data acquisition of covid-19 in six continents through PySpark - a big data tool. medRxiv, p. 2021.07.04.21259983, Jul. 06, 2021. doi: 10.1101/2021.07.04.21259983.
• A. Gupta, H. K. Thakur, R. Shrivastava, P. Kumar, and S. Nag, A big data analysis framework using apache spark and deep learning, in 2017 IEEE International Conference on Data Mining Workshops (ICDMW), Nov. 2017, pp. 9–16. doi: 10.1109/ICDMW.2017.9.
• K. Rothauge, H. Ayyalasomayajula, K. J. Maschhoff, M. Ringenburg, and M. W. Mahoney, Running alchemist on cray xc and cs series supercomputers: dask and PySpark interfaces, deployment options, and data transfer times. arXiv, Nov. 28, 2019. doi: 10.48550/arXiv.1910.01354.
• S. Gafner et al., Evaluation of hip abductor and adductor strength in the elderly: a reliability study, Eur. Rev. Aging Phys. Act., vol. 14, no. 1, p. 5, Apr. 2017, doi: 10.1186/s11556-017-0174-6.
• A. Abraham et al., Machine learning for neuroimaging with Scikit-Learn, Front. Neuroinformatics, vol. 8, 2014, doi: 10.3389/fninf.2014.00014.
• L. Buitinck et al., API design for machine learning software: experiences from the Scikit-Learn project. arXiv, Sep. 01, 2013. doi: 10.48550/arXiv.1309.0238.
• A. Auti, D. Patil, O. Zagade, P. Bhosale, and P. Ahire, Bitcoin price prediction using svm, vol. 6, no. 11, 2022.
• J. Hao and T. K. Ho, Machine learning made easy: a review of Scikit-Learn package in python programming language, J. Educ. Behav. Stat., vol. 44, no. 3, pp. 348–361, Jun. 2019, doi: 10.3102/1076998619832248.
• M. W. Liemohn et al., Model evaluation guidelines for geomagnetic index predictions, Space Weather, vol. 16, no. 12, pp. 2079–2102, 2018, doi: 10.1029/2018SW002067.
• E. Uzunhisarcıklı, E. Kavuncuoğlu, and A. T. Özdemir, Investigating classification performance of hybrid deep learning and machine learning architectures on activity recognition, Comput. Intell., vol. 38, no. 4, Art. no. 4, 2022, doi: 10.1111/coin.12517.
• E. Kavuncuoğlu, E. Uzunhisarcıklı, B. Barshan, and A. T. Özdemir, Investigating the performance of wearable motion sensors on recognizing falls and daily activities via machine learning, Digit. Signal Process., p. 103365, Dec. 2021, doi: 10.1016/J.DSP.2021.103365.
• M. Ş. Turan and B. Barshan, Classification of fall directions via wearable motion sensors, Digit. Signal Process., p. 103129, Jun. 2021, doi: 10.1016/j.dsp.2021.103129.
• A. T. Özdemir, An analysis on sensor locations of the human body for wearable fall detection devices: Principles and practice, Sens. Switz., vol. 16, no. 8, Art. no. 8, Jul. 2016, doi: 10.3390/s16081161.
• D. C. Montgomery, E. A. Peck, and G. G. Vining, Introduction to linear regression analysis. John Wiley & Sons, 2013.
• J. S. Bendat and A. G. Piersol, Random data: analysis and measurement procedures. John Wiley & Sons, 2011.
• J. L. Devore, Probability and statistics for engineering and the sciences. Cengage Learning, 2015.
• N. L. Johnson, S. Kotz, and N. Balakrishnan, Continuous univariate distributions, Vol. 1, 2nd edition. New York: Wiley-Interscience, 1994.
• A. V. Oppenheim and R. W. Schafer, Discrete-time signal processing. Pearson, 2010.
• J. G. Proakis and D. G. Manolakis, Digital signal processing: principles, algorithms, and applications. Macmillan, 1992.
• Signal processing, in biomechanics and motor control of human movement, John Wiley & Sons, Ltd, 2009, pp. 14–44. doi: 10.1002/9780470549148.ch2.
• P. J. Brockwell and R. A. Davis, Introduction to time series and forecasting. in Springer Texts in Statistics. Cham: Springer International Publishing, 2016. doi: 10.1007/978-3-319-29854-2.
• T. Inouye et al., Quantification of eeg irregularity by use of the entropy of the power spectrum, Electroencephalogr. Clin. Neurophysiol., vol. 79, no. 3, pp. 204–210, Sep. 1991, doi: 10.1016/0013-4694(91)90138-T.
• G. H. Golub and C. F. V. Loan, Matrix computations. JHU Press, 2013.
• A. Mannini and A. M. Sabatini, Machine learning methods for classifying human physical activity from on-body accelerometers, Sensors, vol. 10, no. 2, Art. no. 2, Feb. 2010, doi: 10.3390/s100201154.
• T. M. Cover and J. A. Thomas, Elements of information theory. John Wiley & Sons, 2012.
• T. Hastie, R. Tibshirani, and J. Friedman, Overview of supervised learning, in the elements of statistical learning: data mining, inference, and prediction, T. Hastie, R. Tibshirani, and J. Friedman, Eds., in Springer Series in Statistics. , New York, NY: Springer, 2009, pp. 9–41. doi: 10.1007/978-0-387-84858-7_2.
• C. J. C. Burges, A tutorial on support vector machines for pattern recognition, Data Min. Knowl. Discov., vol. 2, no. 2, pp. 121–167, Jun. 1998, doi: 10.1023/A:1009715923555.
• H. T. Babacan, Ö. Yüksek, and F. Saka, Yapay zeka ve sezgisel regresyon yöntemlerinin yağış-akış modellemesi için performans değerlendirmesi: Aksu Deresi için bir uygulama, Niğde Ömer Halisdemir Üniversitesi Mühendis. Bilim. Derg., vol. 11, no. 3, Art. no. 3, Jul. 2022, doi: 10.28948/ngumuh.1079616.
• L. Breiman, Random forests, Mach. Learn., vol. 45, no. 1, pp. 5–32, Oct. 2001, doi: 10.1023/A:1010933404324.
• J. H. Friedman, Greedy function approximation: a gradient boosting machine, 2001.
• Introduction to the logistic regression model, in Applied Logistic Regression, John Wiley & Sons, Ltd, 2013, pp. 1–33. doi: 10.1002/9781118548387.ch1.
• M. Demi̇rhan and S. Behdi̇oğlu, Sağlık çalışanlarının maruz kaldığı şiddetin sıralı lojistik regresyon analizi ile incelenmesi, Toplum Ekon. Ve Önetim Derg., vol. 4, no. 1, Art. no. 1, Jun. 2023, doi: 10.58702/teyd.1228283.
• J. Shi, D. Chen, and M. Wang, Pre-impact fall detection with cnn-based class activation mapping method, Sensors, vol. 20, no. 17, Art. no. 17, Jan. 2020, doi: 10.3390/s20174750.
• S. Zulj, G. Seketa, I. Lackovic, and R. Magjarevic, Accuracy comparison of ml-based fall detection algorithms using two different acceleration derived feature vectors, in World Congress on Medical Physics and Biomedical Engineering 2018, L. Lhotska, L. Sukupova, I. Lacković, and G. S. Ibbott, Eds., in IFMBE Proceedings. Singapore: Springer, 2019, pp. 481–485. doi: 10.1007/978-981-10-9038-7_89.
• C. M. Lee, J. Park, S. Park, and C. H. Kim, Fall-detection algorithm using plantar pressure and acceleration data, Int. J. Precis. Eng. Manuf., vol. 21, no. 4, pp. 725–737, Apr. 2020, doi: 10.1007/s12541-019-00268-w.
• H. Gjoreski et al., Wearable sensors data-fusion and machine-learning method for fall detection and activity recognition, Stud. Syst. Decis. Control, vol. 273, pp. 81–96, 2020, doi: 10.1007/978-3-030-38748-8_4.
• O. Ojetola, E. I. Gaura, and J. Brusey, Fall detection with wearable sensors–safe (smart fall detection), in 2011 Seventh International Conference on Intelligent Environments, Jul. 2011, pp. 318–321. doi: 10.1109/IE.2011.38.