Academia.eduAcademia.edu

Figure 1 – uploaded by Talitha Amory

See full PDFdownloadDownload figure
DNA MICROARRAY DATASET SPECIFICATIONS [4] The preprocessing process corrects problems that arise in the data processing. Such as data with too many features and the high difference in range values in each feature. These problems can cause the results of data processing to be less good or not optimal. The data used in this final project is DNA microarray data from Kent Ridge Biomedical Dataset [4]. The DNA microarray data consists of ovarian cancer, lung cancer, breast cancer, and colon cancer. The dataset specifications used can be seen in Table I.

Table 1 DNA MICROARRAY DATASET SPECIFICATIONS [4] The preprocessing process corrects problems that arise in the data processing. Such as data with too many features and the high difference in range values in each feature. These problems can cause the results of data processing to be less good or not optimal. The data used in this final project is DNA microarray data from Kent Ridge Biomedical Dataset [4]. The DNA microarray data consists of ovarian cancer, lung cancer, breast cancer, and colon cancer. The dataset specifications used can be seen in Table I.

Related Figures (6)

PRECISION RESULTS OF MICROARRAY DATA USING SVM-RFE AND NAIVE BAYES ACCURACY RESULTS OF MICROARRAY DATA USING SVM-RFE AND NAIVE BAYES TABLE II
PRECISION RESULTS OF MICROARRAY DATA USING SVM-RFE AND NAIVE BAYES ACCURACY RESULTS OF MICROARRAY DATA USING SVM-RFE AND NAIVE BAYES TABLE II
The accuracy of the DNA microarray data using the SVM-RFE and Naive Bayes methods is shown in Table II. Based on Table II, it was found that the lung cancer and ovarian cancer dataset obtained better accuracy values compared to other datasets on each number of features. An accuracy value of 100% is obtained by lung cancer data using 50, 100, and 200 features. In the ovarian cancer data, the greatest accuracy value was 99.6% obtained when using ten features. Colon cancer data gets the highest accuracy value of 90% when using ter features. In the breast cancer data, the greatest accuracy value is 93.7% obtained when using 50 features. The highest precision results are also owned by lung cancer data with a precision level of up to 100%. Ovariar annar data hac a atahla neanioinn wala fenm thea noe nf © faativac tr INND factieac whinh to INGO. The hac F1- SCORE RESULTS OF MICROARRAY DATA USING SVM-RFE AND NAIVE BAYES TABLE V
The accuracy of the DNA microarray data using the SVM-RFE and Naive Bayes methods is shown in Table II. Based on Table II, it was found that the lung cancer and ovarian cancer dataset obtained better accuracy values compared to other datasets on each number of features. An accuracy value of 100% is obtained by lung cancer data using 50, 100, and 200 features. In the ovarian cancer data, the greatest accuracy value was 99.6% obtained when using ten features. Colon cancer data gets the highest accuracy value of 90% when using ter features. In the breast cancer data, the greatest accuracy value is 93.7% obtained when using 50 features. The highest precision results are also owned by lung cancer data with a precision level of up to 100%. Ovariar annar data hac a atahla neanioinn wala fenm thea noe nf © faativac tr INND factieac whinh to INGO. The hac F1- SCORE RESULTS OF MICROARRAY DATA USING SVM-RFE AND NAIVE BAYES TABLE V
LALITHA KAYLA AMORY ET. AL./ J. DATA SCI. APPL. 2U2U, 5 (2): 46-90 Comparative Analysis of Support Vector Machine-Recursive Feature Elimination and Chi-Square on Microarray Classification for Cancer Detection with Naive Bayes 5 The accuracy of the microarray DNA data using the Chi-Square and Naive Bayes methods is shown in Table VI. Based on Table VI, it is found that the lung cancer dataset gets a better accuracy value compared to other datasets on each feature count test. An accuracy value of 100% is obtained by lung cancer data when using 200 features. In the ovarian cancer data, the greatest accuracy value is 96.9% obtained when using 100 features. Colon cancer data gets the highest accuracy score of 86.7% when using ten features. However, the model still does not provide good accuracy in breast cancer data. The Chi-Square Feature Selection cannot handle breast cancer data where this data has the highest data complexity compared to other data.
LALITHA KAYLA AMORY ET. AL./ J. DATA SCI. APPL. 2U2U, 5 (2): 46-90 Comparative Analysis of Support Vector Machine-Recursive Feature Elimination and Chi-Square on Microarray Classification for Cancer Detection with Naive Bayes 5 The accuracy of the microarray DNA data using the Chi-Square and Naive Bayes methods is shown in Table VI. Based on Table VI, it is found that the lung cancer dataset gets a better accuracy value compared to other datasets on each feature count test. An accuracy value of 100% is obtained by lung cancer data when using 200 features. In the ovarian cancer data, the greatest accuracy value is 96.9% obtained when using 100 features. Colon cancer data gets the highest accuracy score of 86.7% when using ten features. However, the model still does not provide good accuracy in breast cancer data. The Chi-Square Feature Selection cannot handle breast cancer data where this data has the highest data complexity compared to other data.
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved