Schiff Bases as Anticorrosive Additives for Mild Steel Corrosion in Acid Media

M. Abirami,S. Sasikala,S. Chitra,K. Parameswari,A. Selvaraj 한국부식방식학회(구 한국부식학회) 2009 Corrosion Science and Technology Vol.8 No.1

Analysis and Identifying of Important Features on Road Accidents by using Machine Learning Algorithms

( Manoj Kushwaha ),( M. S. Abirami ),( Corresponding Author M. S. Abirami ) 한국감성과학회 2021 추계학술대회 Vol.2021 No.0

Accidents occurred usually on roads, which bring enormous losses to society. Road accidents are a universal problem which causes the loss of precious human lives and property. The purpose of this paper is to extract important influence features of road accidents and reduce the dimensionality of datasets for getting better results from machine learning algorithms. Collected datasets from Kaggle and constructed new datasets from existing datasets based on the influence feature of road accidents and perform preprocessing, feature selection and feature extraction. Feature selection is done using heat map and correlation matrix. Feature extraction is done using dimensionality reduction methods such as the Principal Component Analysis (PCA), Linear discriminate analysis (LDA) and t-Distributed Stochastic Neighbor Embedding (t-SNE). The different feature extraction techniques are applied and the results are compared based on the accuracy parameter. It was found that LDA performs better than PCA with accuracy of 85% which uses Random Forest classifier.

Analysis and Identifying of Important Features on Road Accidents by using Machine Learning Algorithms

( Manoj Kushwaha ),( M. S. Abirami ) 한국감성과학회 2021 한국감성과학회 국제학술대회(ICES) Vol.2021 No.-

Accidents occurred usually on roads, which bring enormous losses to society. Road accidents are a universal problem which causes the loss of precious human lives and property. The purpose of this paper is to extract important influence features of road accidents and reduce the dimensionality of datasets for getting better results from machine learning algorithms. Collected datasets from Kaggle and constructed new datasets from existing datasets based on the influence feature of road accidents and perform preprocessing, feature selection and feature extraction. Feature selection is done using heat map and correlation matrix. Feature extraction is done using dimensionality reduction methods such as the Principal Component Analysis (PCA), Linear discriminate analysis (LDA) and t-Distributed Stochastic Neighbor Embedding (t-SNE). The different feature extraction techniques are applied and the results are compared based on the accuracy parameter. It was found that LDA performs better than PCA with accuracy of 85% which uses Random Forest classifier.

Analysis and Comparative study on Maternal Mortality for Diabetes prediction using different Machine Learning Algorithms

( G. Keerthi ),( M. S. Abirami ) 한국감성과학회 2021 한국감성과학회 국제학술대회(ICES) Vol.2021 No.-

Diabetes is a leading reason of death, disability, and economic loss around the world. Type 2 diabetes is the maximum shared kind of diabetes in women (80-90 percent worldwide).It can be avoided or postponed by receiving the appropriate maintenance and interventions, including an initial diagnosis. There has remained a lot of progress in the area of medical diagnosis using many machine learning algorithms. However, due to incomplete medical data sets, accuracy suffers, resulting in a higher frequency of misclassifications, which might lead to dangerous complications. Many researchers find that accurately predicting and diagnosing a disease is a difficult scientific topic. As a result, the goal was to improve the diagnostic. The first technique is to collect the dataset, which comprises of 769 pregnant women's records. On the foundation of accuracy, machine learning approaches are utilized to forecast diabetes and non-diabetes women. We used seven machine learning algorithms to calculate diabetes using the dataset. We discovered that a diabetes prediction model that combines Linear Regression and Support Vector Machine performs well, with an accuracy of 77 percent -78 percent.

Analysis and Comparative study on Maternal Mortality for Diabetes prediction using different Machine Learning Algorithms

( G. Keerthi ),( Dr. M. S. Abirami ) 한국감성과학회 2021 추계학술대회 Vol.2021 No.0

Diabetes is a leading reason of death, disability, and economic loss around the world. Type 2 diabetes is the maximum shared kind of diabetes in women (80-90 percent worldwide).It can be avoided or postponed by receiving the appropriate maintenance and interventions, including an initial diagnosis. There has remained a lot of progress in the area of medical diagnosis using many machine learning algorithms. However, due to incomplete medical data sets, accuracy suffers, resulting in a higher frequency of misclassifications, which might lead to dangerous complications. Many researchers find that accurately predicting and diagnosing a disease is a difficult scientific topic. As a result, the goal was to improve the diagnostic. The first technique is to collect the dataset, which comprises of 769 pregnant women's records. On the foundation of accuracy, machine learning approaches are utilized to forecast diabetes and non-diabetes women. We used seven machine learning algorithms to calculate diabetes using the dataset. We discovered that a diabetes prediction model that combines Linear Regression and Support Vector Machine performs well, with an accuracy of 77 percent -78 percent.

Analysis of Various Machine Learning Algorithms for the Prediction of Heart Disease

( Poomari Durga. K ),( M. S. Abirami ) 한국감성과학회 2021 한국감성과학회 국제학술대회(ICES) Vol.2021 No.-

cardiovascular disease is one of the major reasons of mortality in the society. Prediction of heart disease is a big challenge in the area of medicinal field. The purpose of this paper is to extract important impact features which predicts the heart disease and reduce the dimensionality of datasets for getting better results from machine learning algorithm. In this paper we propose a IOT based Remote Cardiac Monitoring Model which transfers the Realtime patient data to a medical service in case of abnormality found in the patient. This paper labels about various attributes, symptoms, datasets related to heart disease prediction. The next step is analysis of various heart diseases and collected various heart disease related dataset. The final stage is to perform pre-processing and feature selection with the help of machine learning Techniques. Various attributes, symptoms and datasets related to prediction of heart disease is analyzed and comparative analysis of accuracy obtained by various machine learning technique to accurately predict the heart disease is done with the evaluation metrics such as RMSE, MAE, precision, recall. Heart attack is unpredictable in medical field which may cause loss of human life. This can be achieved with the machine learning techniques and provide sudden treatment and save the life before going critical. Most of the death cases due to severe heart attack. With the help of wearable IOT devices we can predict the abnormal changes in heart beat and send to nearby medical unit and save the life

Analysis of Various Machine Learning Algorithms for the Prediction of Heart Disease

( Poomari Durga. K ),( M. S. Abirami ) 한국감성과학회 2021 추계학술대회 Vol.2021 No.0

cardiovascular disease is one of the major reasons of mortality in the society. Prediction of heart disease is a big challenge in the area of medicinal field. The purpose of this paper is to extract important impact features which predicts the heart disease and reduce the dimensionality of datasets for getting better results from machine learning algorithm. In this paper we propose a IOT based Remote Cardiac Monitoring Model which transfers the Realtime patient data to a medical service in case of abnormality found in the patient. This paper labels about various attributes, symptoms, datasets related to heart disease prediction. The next step is analysis of various heart diseases and collected various heart disease related dataset. The final stage is to perform pre-processing and feature selection with the help of machine learning Techniques. Various attributes, symptoms and datasets related to prediction of heart disease is analyzed and comparative analysis of accuracy obtained by various machine learning technique to accurately predict the heart disease is done with the evaluation metrics such as RMSE, MAE, precision, recall. Heart attack is unpredictable in medical field which may cause loss of human life. This can be achieved with the machine learning techniques and provide sudden treatment and save the life before going critical. Most of the death cases due to severe heart attack. With the help of wearable IOT devices we can predict the abnormal changes in heart beat and send to nearby medical unit and save the life