Search

Predicting the onset of cardiovascular disease is the goal. Images are separated into labelled and unlabeled categories using two machine learning algorithms: Decision Tree and Logistic regression. In order to efficiently and accurately analyse labelled pictures with G power in 80% and threshold 0.05 percent, CI 95% mean and standard deviation, the sample size was iterated 10 times from n =5 in Decision Tree and n = 5 in Logistic Regression. Accurate predictions and classifications of values from cardiac patient data have been generated in this study by comparing the predictive and classifying abilities of Decision Tree and Logistic Regression. The accuracy of Decision Tree (72.40%) is statistically significant (p0.05) greater than that of Logistic Regression (67.59%). Decision Tree outperforms Logistic Regression in predicting cardiovascular disease. [ABSTRACT FROM AUTHOR]

The goal is to anticipate cardiac arrest in heart disease patients. Machine learning methods like SVM and decision tree categorise photos (DT). To effectively and reliably analyse labelled pictures with G power of 80%, threshold 0.05 percent, CI 95 percent mean and standard deviation, SVM and Decision Tree sample sizes of n=5 were iterated 10 times. The Support Vector Machine (SVM) outperformed the Decision Tree (DT) in predicting and categorising cardiac patients' data with a p-value of 0.05. Support Vector Machine predicts cardiomyopathy risk better than Decision Tree. [ABSTRACT FROM AUTHOR]

Signal processing and multimedia are among the best applications for imprecise computing. The use of precise computation in arithmetic circuits can deliver more useful results without consuming more electricity. Using recursive multiplication by approximate multipliers in the form of compressors, we propose a new design that takes advantage of the partitions of partial products. An approximate compression of 4:2 is used to propose three multiplier designs. Compared with previous approximate multiplier designs, the proposed design demonstrated makes significant improvement in accuracy as well as power and area reductions. The approximate compressor is used for the first two multiplier designs. With a reduced power, area, and accuracy compared with existing multipliers, Xilinx simulates and synthesizes the proposed multiplier design (1) in a Xilinx simulation. Xilinx software is also used to simulate and synthesize the proposed multiplier design (2) with high accuracy. The proposed multiplier design (3) makes use of a second approximate compressor. In comparison to designs (1) and (2), it consumes less power and area. [ABSTRACT FROM AUTHOR]