Performance Analysis and Comparison of Machine and Deep Learning Algorithms for IoT Data Classification

Meysam Vakili,Mohammad Ghamsari,Masoumeh Rezaei

DOI: https://doi.org/10.48550/arXiv.2001.09636

2020-01-27

Abstract:In recent years, the growth of Internet of Things (IoT) as an emerging technology has been unbelievable. The number of networkenabled devices in IoT domains is increasing dramatically, leading to the massive production of electronic data. These data contain valuable information which can be used in various areas, such as science, industry, business and even social life. To extract and analyze this information and make IoT systems smart, the only choice is entering artificial intelligence (AI) world and leveraging the power of machine learning and deep learning techniques. This paper evaluates the performance of 11 popular machine and deep learning algorithms for classification task using six IoT-related datasets. These algorithms are compared according to several performance evaluation metrics including precision, recall, f1-score, accuracy, execution time, ROC-AUC score and confusion matrix. A specific experiment is also conducted to assess the convergence speed of developed models. The comprehensive experiments indicated that, considering all performance metrics, Random Forests performed better than other machine learning models, while among deep learning models, ANN and CNN achieved more interesting results.

Machine Learning

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how different machine learning and deep learning algorithms perform in the Internet of Things (IoT) data classification tasks. With the development of IoT technology, the number of devices connected to the Internet has increased dramatically, resulting in the generation of a large amount of electronic data. These data contain valuable information and can be used in multiple fields such as science, industry, commerce, and even social life. In order to extract and analyze information from these data and make IoT systems more intelligent and enter the world of artificial intelligence, it has become an inevitable choice to utilize machine learning and deep learning technologies.

The paper specifically studied the performance of 11 popular machine learning and deep learning algorithms in the classification tasks on six IoT - related datasets. These algorithms were compared through multiple performance evaluation metrics (such as precision, recall, F1 - score, accuracy, execution time, ROC - AUC score, and confusion matrix). In addition, a specific experiment was also carried out to evaluate the convergence speed of the developed models.

The main purpose of the research is to provide a comprehensive study to analyze and compare the performance of several common machine learning and deep learning algorithms, with particular attention to the classification problem, because this is very important in a large number of IoT applications. Through this study, the author hopes to provide better knowledge and insights for researchers and practitioners in related fields so that they can choose appropriate algorithms according to the problems they face.