Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9640928
M. Surendar, P. Pradeepa
Battery surveillance is critical for the majority of battery-powered vehicles, for the benefit of the lead-acid battery's safety , functioning, and even to extend its life. Due to the development of EVs and HEVs, battery technology has made tremendous progress in recent years. However, the estimate of the state of charge (SOC) remains a battery engineering challenge. The remaining load ratio to the maximum load battery capacity is defined as the SOC. In terms of battery safety and maintenance, the SOC estimate is of prime importance. Artificial intelligence, notably machine learning-based systems, has recently been used to estimate battery state, both as part of adaptive systems and as stand-alone systems. The use of data-driven algorithms to estimate battery conditions with high precision is a potential approach. The purpose of this study is to offer a novel and highly accurate approach for predicting the state of charge (SOC) of a Li-ion battery cell that requires little conceptualization and modeling work. The battery aging process can be slowed down by properly treating the battery, including restricting frequent charge and deep drain cycles. This study presents an analysis based on IoT with an ultimate wireless battery surveillance system (WBSS) to determine the relationship between journey distance and discharge cycle. The proposed system's methodology has been tested and found to be effective.
{"title":"An IOT-based Battery Surveillance System For E-Vehicles","authors":"M. Surendar, P. Pradeepa","doi":"10.1109/I-SMAC52330.2021.9640928","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9640928","url":null,"abstract":"Battery surveillance is critical for the majority of battery-powered vehicles, for the benefit of the lead-acid battery's safety , functioning, and even to extend its life. Due to the development of EVs and HEVs, battery technology has made tremendous progress in recent years. However, the estimate of the state of charge (SOC) remains a battery engineering challenge. The remaining load ratio to the maximum load battery capacity is defined as the SOC. In terms of battery safety and maintenance, the SOC estimate is of prime importance. Artificial intelligence, notably machine learning-based systems, has recently been used to estimate battery state, both as part of adaptive systems and as stand-alone systems. The use of data-driven algorithms to estimate battery conditions with high precision is a potential approach. The purpose of this study is to offer a novel and highly accurate approach for predicting the state of charge (SOC) of a Li-ion battery cell that requires little conceptualization and modeling work. The battery aging process can be slowed down by properly treating the battery, including restricting frequent charge and deep drain cycles. This study presents an analysis based on IoT with an ultimate wireless battery surveillance system (WBSS) to determine the relationship between journey distance and discharge cycle. The proposed system's methodology has been tested and found to be effective.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128593086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9640765
K. Karthik, S. Rajaprakash, S. Nazeeb Ahmed, Rishan Perincheeri, C. R. Alexander
The main challenge to the farmers is that the weather, environmental factors cannot be predicted and controlled. Plant diseases also plays an important role in plant cultivation. Plant diseases are considered to be a major challenge to the farmers. As plant and leaf diseases is difficult to be identified with the naked eyes. To overcome this issue in the existing approach, the farmers periodically spray pesticides which might spoil the plants, crop failure. Thus, effective monitoring and identification of plant leaf disease detection at the early stage is essential to predict the leaf diseases and recommend preventive measures. The proposed system utilizes image processing with deep learning techniques to detect plant leaf diseases from potato and tomato datasets. Also, our proposed system could able to recommend the plant benefits helping the current generation of people with a common knowledge base along with plant leaf diseases prediction. For experimental results, this research uses jupyter tool with python script for performing plant leaf disease analysis.
{"title":"Tomato And Potato Leaf Disease Prediction With Health Benefits Using Deep Learning Techniques","authors":"K. Karthik, S. Rajaprakash, S. Nazeeb Ahmed, Rishan Perincheeri, C. R. Alexander","doi":"10.1109/I-SMAC52330.2021.9640765","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9640765","url":null,"abstract":"The main challenge to the farmers is that the weather, environmental factors cannot be predicted and controlled. Plant diseases also plays an important role in plant cultivation. Plant diseases are considered to be a major challenge to the farmers. As plant and leaf diseases is difficult to be identified with the naked eyes. To overcome this issue in the existing approach, the farmers periodically spray pesticides which might spoil the plants, crop failure. Thus, effective monitoring and identification of plant leaf disease detection at the early stage is essential to predict the leaf diseases and recommend preventive measures. The proposed system utilizes image processing with deep learning techniques to detect plant leaf diseases from potato and tomato datasets. Also, our proposed system could able to recommend the plant benefits helping the current generation of people with a common knowledge base along with plant leaf diseases prediction. For experimental results, this research uses jupyter tool with python script for performing plant leaf disease analysis.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130044281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9640699
Jianzhang Zhao, Jing Shen
Textile products are social, scientific and natural. Consumers' choice of goods is often affected by mental state, emotional attributes, life background and other factors. The advantage of printing and chemical technology and the increasing innovation of textile equipment provide a good prerequisite for designers to choose colors and achieve the desired color display effect. Digital printing technology is a new type of printing technology integrating electronic information, computers, machinery and other organs, provides a new method to study the shape and structure of different entities, and also provides a theoretical basis for the emergence of fractal art. Aiming at the defects of single mathematical model and few types of fractal graphics, this paper put out a fractal graphics generation way based on self combination nonlinear transformation. Based on the source code of apophysis, this paper develops a variety of new adaptive models, and achieves batch generation of fractal art graphics through self combination model.
{"title":"Application of Fractal Geometry in Textile Digital Printing Pattern Design","authors":"Jianzhang Zhao, Jing Shen","doi":"10.1109/I-SMAC52330.2021.9640699","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9640699","url":null,"abstract":"Textile products are social, scientific and natural. Consumers' choice of goods is often affected by mental state, emotional attributes, life background and other factors. The advantage of printing and chemical technology and the increasing innovation of textile equipment provide a good prerequisite for designers to choose colors and achieve the desired color display effect. Digital printing technology is a new type of printing technology integrating electronic information, computers, machinery and other organs, provides a new method to study the shape and structure of different entities, and also provides a theoretical basis for the emergence of fractal art. Aiming at the defects of single mathematical model and few types of fractal graphics, this paper put out a fractal graphics generation way based on self combination nonlinear transformation. Based on the source code of apophysis, this paper develops a variety of new adaptive models, and achieves batch generation of fractal art graphics through self combination model.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127785503","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9640900
Kaushal Mehta, Sachin Sharma, Dipankar Mishra
Fire, as one of the world's biggest calamities, must be identified at the right moment before it can do significant damage to the atmosphere and living beings. According to a study, 75-80 percent of the various casualties caused by fire might have been prevented if the misfortune was understood quickly. Particularly in the case of a forest fire, this results in a significant loss to the environment and makes it extremely dangerous for animals to remain there. To avoid such losses, an automated system is needed that can provide early detection of any fire situation via any of the alarm systems. This paper examines the IoT's momentum, advances, and applications in the fire-fighting industry. In addition, the paper summarises a survey conducted for identifying research trends and difficulties in fire projects. The fire Internet of Things (IoT) aims to link different objects with organisations in the fire domain. This paper describes the creation of a fire detector using Arduino, which is equipped with smoke and temperature sensors and emits a buzzer alarm in response to the findings.
{"title":"Internet-of-Things Enabled Forest Fire Detection System","authors":"Kaushal Mehta, Sachin Sharma, Dipankar Mishra","doi":"10.1109/I-SMAC52330.2021.9640900","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9640900","url":null,"abstract":"Fire, as one of the world's biggest calamities, must be identified at the right moment before it can do significant damage to the atmosphere and living beings. According to a study, 75-80 percent of the various casualties caused by fire might have been prevented if the misfortune was understood quickly. Particularly in the case of a forest fire, this results in a significant loss to the environment and makes it extremely dangerous for animals to remain there. To avoid such losses, an automated system is needed that can provide early detection of any fire situation via any of the alarm systems. This paper examines the IoT's momentum, advances, and applications in the fire-fighting industry. In addition, the paper summarises a survey conducted for identifying research trends and difficulties in fire projects. The fire Internet of Things (IoT) aims to link different objects with organisations in the fire domain. This paper describes the creation of a fire detector using Arduino, which is equipped with smoke and temperature sensors and emits a buzzer alarm in response to the findings.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127892176","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9641053
Paras Jain, M. Murali, Amaan Ali
Facial emotion recognition is an emerging research field in detecting Facial Expression. Deep learning algorithms have gained immense success in different areas of implementation such as classification, recommendation models, object recognition etc. The various types of modules that are brought together in this technique for the betterment of the working of the model is mainly contributed by the progress in the field of Deep Learning. The main focus of this work is to create a Neural Network model which is capable of classifying human emotions in a set of 7 different classes. Image data is used for testing, validation, and training of the model.
{"title":"Face Emotion Detection Using Deep Learning","authors":"Paras Jain, M. Murali, Amaan Ali","doi":"10.1109/I-SMAC52330.2021.9641053","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9641053","url":null,"abstract":"Facial emotion recognition is an emerging research field in detecting Facial Expression. Deep learning algorithms have gained immense success in different areas of implementation such as classification, recommendation models, object recognition etc. The various types of modules that are brought together in this technique for the betterment of the working of the model is mainly contributed by the progress in the field of Deep Learning. The main focus of this work is to create a Neural Network model which is capable of classifying human emotions in a set of 7 different classes. Image data is used for testing, validation, and training of the model.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128300252","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9641004
K. Karthik, S. Rajaprakash, Mohammad Adil Aboobacker, T. A. Backer, Sajan K Davis
In this digital era, data plays an indispensible role in human lives. In particular, data is more available in different platforms such has social media, healthcare, education etc. Most of the time, the digital data are prone to the cyber attacks. To overcome this challenge, this research work applies the novel technique called RB01. This technique has four main stages. The first stage applies the T-test technique; the second stage applies the odd operations; the third stage applies the even operations; and the last stage applies the quadratic equations. The proposed RB01 technique leverages high security while compared to ChaCha method.
{"title":"RB01 Technique Used To Applying The Generalized Data","authors":"K. Karthik, S. Rajaprakash, Mohammad Adil Aboobacker, T. A. Backer, Sajan K Davis","doi":"10.1109/I-SMAC52330.2021.9641004","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9641004","url":null,"abstract":"In this digital era, data plays an indispensible role in human lives. In particular, data is more available in different platforms such has social media, healthcare, education etc. Most of the time, the digital data are prone to the cyber attacks. To overcome this challenge, this research work applies the novel technique called RB01. This technique has four main stages. The first stage applies the T-test technique; the second stage applies the odd operations; the third stage applies the even operations; and the last stage applies the quadratic equations. The proposed RB01 technique leverages high security while compared to ChaCha method.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126328363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9640994
A. Christy, M. A. Anto Praveena, S. Vaithyasubramanian, M. Roobini
In today’s world Chemical industries and biochemical plants deal with hazardous chemical reactions. The progression of the reaction / completion of the reaction are done by observation of different parameters as well as aliquoting different samples for analysis. This normal procedure is tedious, time consuming and cumbersome. A novel method has been developed to accurately estimate the end point of chemical reaction in remote locations using Internet of Things (IoT) platform. The procedure involves monitoring different parameters like changes in pressure, volume and analyzing the data to find accurately the end point of locations using IOT and Machine learning techniques. A simple reaction involving oxidation of oxalic acid with incremental addition of potassium permanganate in nitric acid medium was carried out in lab scale under vacuum and the drop in vacuum was observed with time and volume. The end point of the reaction was accurately estimated by observing the pressure values using a pressure sensor and passing it to the cloud through the Wi-Fi module. Data is analyzed using machine learning techniques and once the curve flattens means the end point is reached. The data sends an alert to the IoT device that the reaction is completed as well as the circuit is automatically stopped in further functioning.
{"title":"Completion of Chemical Reaction in Remote Locations Using Data Analytics Built on Internet of Things Platform","authors":"A. Christy, M. A. Anto Praveena, S. Vaithyasubramanian, M. Roobini","doi":"10.1109/I-SMAC52330.2021.9640994","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9640994","url":null,"abstract":"In today’s world Chemical industries and biochemical plants deal with hazardous chemical reactions. The progression of the reaction / completion of the reaction are done by observation of different parameters as well as aliquoting different samples for analysis. This normal procedure is tedious, time consuming and cumbersome. A novel method has been developed to accurately estimate the end point of chemical reaction in remote locations using Internet of Things (IoT) platform. The procedure involves monitoring different parameters like changes in pressure, volume and analyzing the data to find accurately the end point of locations using IOT and Machine learning techniques. A simple reaction involving oxidation of oxalic acid with incremental addition of potassium permanganate in nitric acid medium was carried out in lab scale under vacuum and the drop in vacuum was observed with time and volume. The end point of the reaction was accurately estimated by observing the pressure values using a pressure sensor and passing it to the cloud through the Wi-Fi module. Data is analyzed using machine learning techniques and once the curve flattens means the end point is reached. The data sends an alert to the IoT device that the reaction is completed as well as the circuit is automatically stopped in further functioning.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127068485","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9640948
K. Karthik, S. Rajaprakash, Vaddepraveenkumar, Avula Gopinath, K. Chandu
Data is more readily available through social media, hospitals, populations, and other places. The importance of digital data in current and future world cannot be overstated, because data is the only factor that determines the survival of human lives in the world. Despite the hype, digital data is growing more vulnerable to hackers due to a lack of effective security. This research work implements KRB01, a new approach to solve this problem. There are four steps to implement the proposed procedure. The first step applies the T-test technique; The second step applies the odd operations; The third step represents even operations; and the final fourth step applies the equations. The KRB01 method gives high security when compared to ChaCha method.
{"title":"KRB01 Method for Securing the Data","authors":"K. Karthik, S. Rajaprakash, Vaddepraveenkumar, Avula Gopinath, K. Chandu","doi":"10.1109/I-SMAC52330.2021.9640948","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9640948","url":null,"abstract":"Data is more readily available through social media, hospitals, populations, and other places. The importance of digital data in current and future world cannot be overstated, because data is the only factor that determines the survival of human lives in the world. Despite the hype, digital data is growing more vulnerable to hackers due to a lack of effective security. This research work implements KRB01, a new approach to solve this problem. There are four steps to implement the proposed procedure. The first step applies the T-test technique; The second step applies the odd operations; The third step represents even operations; and the final fourth step applies the equations. The KRB01 method gives high security when compared to ChaCha method.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127229116","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9640903
B. Vaishnavi, Borra Harsha, N. N. Chandana, Vemula Bhargavi, Suneetha Manne
Currently, many organizations started using conferencing apps to connect students and teachers, also to conduct online classes. In the current pandemic, applications like google meet, zoom became a necessity for educational institutions to conduct online lectures. An educational institute may need a customized video conferencing system for hassle free online classes, summary of classes and discussion forums. This research work develops a web application with mixed features of video conferencing and report generation. Students who miss their classes can see the reports/summary related to the classes conducted on a particular day and can learn easily. Teachers can login through, create links for their respective classes and share it with their students. Every user utilize their respective login id. Discussion forums are also provided for students to discuss among their peers. Students obtaining quality education is needed which shapes their future. The summary of the classes will help them to learn more along with the lectures. Inter department students can communicate in the discussion forums and support each other.
{"title":"Online Video Conferencing with Report Generation","authors":"B. Vaishnavi, Borra Harsha, N. N. Chandana, Vemula Bhargavi, Suneetha Manne","doi":"10.1109/I-SMAC52330.2021.9640903","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9640903","url":null,"abstract":"Currently, many organizations started using conferencing apps to connect students and teachers, also to conduct online classes. In the current pandemic, applications like google meet, zoom became a necessity for educational institutions to conduct online lectures. An educational institute may need a customized video conferencing system for hassle free online classes, summary of classes and discussion forums. This research work develops a web application with mixed features of video conferencing and report generation. Students who miss their classes can see the reports/summary related to the classes conducted on a particular day and can learn easily. Teachers can login through, create links for their respective classes and share it with their students. Every user utilize their respective login id. Discussion forums are also provided for students to discuss among their peers. Students obtaining quality education is needed which shapes their future. The summary of the classes will help them to learn more along with the lectures. Inter department students can communicate in the discussion forums and support each other.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"173 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127232693","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-11DOI: 10.1109/I-SMAC52330.2021.9640967
K. P. N. V. Satya Sree, J. Karthik, Chava Niharika, P. Srinivas, N. Ravinder, Chitturi Prasad
While some data have an explicit, numerical form, many other data, such as gender or nationality, do not typically use numbers and are referred to as categorical data. Thus, machine learning algorithms need a way of representing categorical information numerically in order to be able to analyze them. Our project specifically focuses on optimizing the conversion of categorical features to a numerical form in order to maximize the effectiveness of various machine learning models. From the methods utilized, it has been observed that wide and deep is the most effective model for datasets that contain high-cardinality features, as opposed to learn embedding and one-hot encoding.
{"title":"Optimized Conversion of Categorical and Numerical Features in Machine Learning Models","authors":"K. P. N. V. Satya Sree, J. Karthik, Chava Niharika, P. Srinivas, N. Ravinder, Chitturi Prasad","doi":"10.1109/I-SMAC52330.2021.9640967","DOIUrl":"https://doi.org/10.1109/I-SMAC52330.2021.9640967","url":null,"abstract":"While some data have an explicit, numerical form, many other data, such as gender or nationality, do not typically use numbers and are referred to as categorical data. Thus, machine learning algorithms need a way of representing categorical information numerically in order to be able to analyze them. Our project specifically focuses on optimizing the conversion of categorical features to a numerical form in order to maximize the effectiveness of various machine learning models. From the methods utilized, it has been observed that wide and deep is the most effective model for datasets that contain high-cardinality features, as opposed to learn embedding and one-hot encoding.","PeriodicalId":178783,"journal":{"name":"2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127563437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：