Classification of Similar Sports Images Using Convolutional Neural Network with Hyper-Parameter Optimization

Podgorelec, Vili; Pečnik, Špela; Vrbančič, Grega

doi:10.3390/app10238494

Cited by 30 publications

(12 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This subsection summarizes machine learning based sport data analysis in the literature. Podgorelec et al [22] built a new image dataset of four similar sports (American football, rugby, soccer, and field hockey) and developed a method to classify those images using transfer learning of CNN with Hyper-Parameter Optimization (HPO). Their proposed method was then compared to a conventional CNN and a CNN with transfer learning but handpicked hyper-parameters for fine-tuning.…”

Section: Machine Learning Based Sport Data Analysismentioning

confidence: 99%

A Dockerized Big Data Architecture for Sports Analytics

Özgüven

Gönener

Eken

2021

Preprint

View full text Add to dashboard Cite

The revolution of big data has also affected the area of sports analytics. Many big companies have started to see the benefits of combining sports analytics and big data to make a profit. Aggregating and processing big sport data from different sources becomes challenging if we rely on central processing techniques, which hurts the accuracy and the timeliness of the information. Distributed systems come to the rescue as a solution to these problems and the MapReduce paradigm is promising for large-scale data analytics. In this study, we present a big data architecture based on Docker containers in Apache Spark. We demonstrate the architecture on four data-intensive case studies including structured analysis, streaming, machine learning methods, and graph-based analysis in sport analytics, showing ease of use.

show abstract

Section: Machine Learning Based Sport Data Analysismentioning

confidence: 99%

A Dockerized Big Data Architecture for Sports Analytics

Özgüven

Gönener

Eken

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…When choosing the best hyper-parameter, it is required to have deep knowledge about machine learning algorithms. The Hyper-parameter optimization technique is using for the automation of selecting the most appropriate parameter [12].…”

Section: Sentiment Image Classification By Convolutional Neural Networkmentioning

confidence: 99%

“…The Term Presence Vs Term Frequency, N-gram Features, Parts of Speech(POS), Term Position, Negation [12], SentiWordNe [43] are several methods that can be used for Twitter data pre-processing and feature extraction.…”

Section: Data Ingestionmentioning

confidence: 99%

Literature review on Real-time Location-Based Sentiment Analysis on Twitter

Rathnayaka¹,

Jayasena²,

Ratnayake³

2021

Adv. Technol.

View full text Add to dashboard Cite

Sentiment analysis mainly supports sorting out the polarity and provides valuable information with the use of raw data in social media platforms. Many fields like health, business, and security require real-time data analysis for instant decision-making situations.Since Twitter is considered a popular social media platform to collect data easily, this paper is considering data analysis methods of Twitter data, real-time Twitter data analysis based on geo-location. Twitter data classification and analysis can be done with the use of diverse algorithms and deciding the most appropriate algorithm for data analysis, can be accomplished by implementing and testing these diverse algorithms.This paper is discussing the major description of sentiment analysis, data collection methods, data pre-processing, feature extraction, and sentiment analysis methods related to Twitter data. Real-time data analysis arises as a major method of analyzing the data available online and the real-time Twitter data analysis process is described throughout this paper. Several methods of classifying the polarized Twitter data are discussed within the paper while depicting a proposed method of Twitter data analyzing algorithm. Location-based Twitter data analysis is another crucial aspect of sentiment analyses, that enables data sorting according to geo-location, and this paper describes the way of analyzing Twitter data based on geo-location. Further, a comparison about several sentiment analysis algorithms used by previous researchers has been reported and finally, a conclusion has been provided.

show abstract

“…Setting the appropriate values of hyper-parameters for the process of training has a direct impact on the final predictive performance of such models, therefore the values should be carefully chosen. While commonly this is still a manual process, a great amount of research was put into developing automatic methods [38,43,49], which would take care of this problem. Since many studies have addressed the problem of identifying a COVID-19 from X-ray images and since the chosen hyper-parameter values have a direct impact on the final classification performance, it is crucial to set hyper-parameter values appropriately especially when addressing such sensitive problem.…”

Section: Introductionmentioning

confidence: 99%

“…Based on our previous experience with the identification of COVID-19 [43], promising results from similar studies [3,36] and our previous work on solving HPO problem [38,43], we set our goal to generalize our GWOTLT [43] from our previous research, in which we utilized the grey wolf optimizer (GWO) algorithm to find the most suitable values of hyper-parameters, to make it agnostic to the usage of different optimization algorithms. Such a generalized HPO method for transfer learning (HPO-TL) enables us to employ various optimization algorithms in order to find the most suitable values of hyper-parameters in order to achieve the best possible predictive model utilizing transfer learning.…”

Section: Introductionmentioning

confidence: 99%

Hyper-parameter optimization of convolutional neural networks for classifying COVID-19 X-ray images

Vrbacic

Pečnik

Podgorelec

2022

ComSIS

Self Cite

View full text Add to dashboard Cite

For more than a year the COVID-19 epidemic is threatening people all over the world. Numerous researchers are looking for all possible insights into the new corona virus SARS-CoV-2. One of the possibilities is an in-depth analysis of X-ray images from COVID-19 patients, commonly conducted by a radiologist, which are due to high demand facing with overload. With the latest achievements in the field of deep learning, the approaches using transfer learning proved to be successful when tackling such problem. However, when utilizing deep learning methods, we are commonly facing the problem of hyper-parameter settings. In this research, we adapted and generalized transfer learning based classification method for detecting COVID-19 from X-ray images and employed different optimization algorithms for solving the task of hyper-parameter settings. Utilizing different optimization algorithms our method was evaluated on a dataset of 1446 X-ray images, with the overall accuracy of 84.44%, outperforming both conventional CNN method as well as the compared baseline transfer learning method. Besides quantitative analysis, we also conducted a qualitative in-depth analysis using the local interpretable modelagnostic explanations method and gain some in-depth view of COVID-19 characteristics and the predictive model perception.

show abstract

Classification of Similar Sports Images Using Convolutional Neural Network with Hyper-Parameter Optimization

Cited by 30 publications

References 57 publications

A Dockerized Big Data Architecture for Sports Analytics

A Dockerized Big Data Architecture for Sports Analytics

Literature review on Real-time Location-Based Sentiment Analysis on Twitter

Hyper-parameter optimization of convolutional neural networks for classifying COVID-19 X-ray images

Contact Info

Product

Resources

About