Special Issue: VOL.-9, Issue-8, August 2023

Abstract

Cardiotocogram (CTG) is one of the noticing mechanical assemblies to evaluate the child prosperity in tummy. CTG generally yields two results fetal prosperity rate (FHR) and uterine choking influences (UC). Through and through, there are 21 attributes in the assessment of FHR and UC on CTG. These attributes can help obstreticians to clasify whether the incipient organism prosperity is run of the mill, thought, or psychotic. This investigation covers the disclosures and assessments of different simulated intelligence models for fetal prosperity game plan. CTG data of 2126 pregnant women were gotten from the School of California Irvine man-made intelligence Storage facility. Ten unique simulated intelligence game plan models were arranged using CTG data. Mindfulness, exactness, and F1 score for each class and as a rule of each model were obtained to expect common, suspect, and masochist fetal states. The data was examined and used in a two ML models. For request, Logistic Regression and Decision Tree form classifier were utilized. Right when the results are investigated, it is found that the Decision Tree rule classifier model conveys the best results. It achieves 97.47% accuracy, which is better than the past strategy declared.

Keywords: Fetal Health Classification Cardiotocogram (CTG) Dataset Decision Tree Classifier Logistic Regression Machine Learning in Prenatal Diagnosis

References

References not available

Abstract

Missing data is a common challenge in the field of medical data analysis, particularly when predicting outcomes from mammographic data. The missing data is one of the typical issues of data quality. An enormous part of the certified datasets have missing characteristics. Crediting the missing characteristics simplifies the assessment by making an all out dataset as it kills the issue of managing complex instances of missingness. In this research paper, we investigate the impact of missing data on the performance of supervised learning models, specifically decision tree classification, when applied to mammographic data. The objective of this examination is to address the impact of missing data on the data mining errand of learning disclosure measure. The principal stage in dealing with the dataset may itself challenge since this improvement requires overseeing missing properties.

Keywords: Mammographic Data Prediction Missing Data Imputation K-Nearest Neighbors (KNN) Imputation Decision Tree Classification Medical Data Preprocessing

References

References not available

Abstract

In multi-name arrangement, every one of the information tests has a place with at least one than one class marks. The customary paired and multi-class order issues are the subset of the multi-name issue with the quantity of marks comparing to each example restricted to one. This exploration researches the use of multiclass arrangement procedures to methodologies to anticipate the result of the medications that may be precise for the patient utilizing Calculated Relapse (LR) calculation with One-Against One (OVO) and One-Versus-Rest (OVR) systems. The exploratory outcomes show the prevalence of LR with OVR accomplishing the most noteworthy exactness of 91.49% with OVO methodology. Additionally, OVR outflanked OVO in LR calculation, exhibiting its viability for multiclass issues. These discoveries offer significant bits of knowledge for drug expectation and further advances the condition of multiclass grouping strategies in computer based intelligence applications.

Keywords: Medication Classification Prediction Multilabel Classification Logistic Regression (LR) One-vs-Rest (OVR) Strategy One-vs-One (OVO) Strategy

References

References not available

Abstract

The prediction of diabetes is a critical task in healthcare, with the potential to significantly improve patient outcomes through early detection and intervention. In this research paper, we conduct a comparative analysis of two machine learning algorithms, Multilayer Perceptron and Naive Bayes, for the prediction of diabetes in the Pima Indian Diabetes dataset. We evaluate the performance of these algorithms in terms of accuracy, precision, and recall, aiming to identify the most effective approach for diabetes prediction.

Keywords: Pima Indian Diabetes Dataset Diabetes Prediction Multilayer Perceptron (MLP) Naïve Bayes Classifier Comparative Machine Learning Analysis

References

References not available