| Abstrakt: |
With the use of deep learning systems like AlexNet and Novel LeNet, the objective of this study is to ascertain the degree to which these systems are able to recognise instances of malaria infection. Methodologies and Instruments for Research: Both the dataset and a number of photos that were used as examples were obtained from the Kaggle website. The Novel LeNet and AlexNet algorithms each have twenty samples gathered for their respective classifications. In conclusion, the sample is comprised of forty individuals. The determination of the total sample size was carried out with the assistance of clinicalc.com, with the following parameters being preserved: The alpha error in this study was 0.5, the threshold value was also 0.5, the confidence level was 95 percent, the G power was 80 percent, and the enrollment ratio was 0.1. For the purpose of determining the accuracy values, Python, Anaconda, and Jupiter were utilised, in addition to a reference dataset. Results: For the purpose of comparing accuracy, an independent sample t-test is carried out with the use of SPSS software. The LeNet technique and the algorithms were used to compare it, and the results showed that it was statistically significant. resulted in successes and achievements. Given that the p-value that was determined is 0.001 (p<0.05), it is possible that the findings might be regarded statistically significant. The finding is that when compared to AlexNet algorithms, unique LeNet algorithms provided greater accuracy values in the detection of malaria illnesses. [ABSTRACT FROM AUTHOR] |
