Protein Interaction Analysis (PPI) can be used to identify proteins that have a supporting function on the main protein, especially in the synthesis process. Insulin is synthesized by proteins that have the same molecular function covering different but mutually supportive roles. To identify this function, the translation of Gene Ontology (GO) gives certain characteristics to each protein. This study purpose to predict proteins that interact with insulin using the centrality method as a feature extractor and extreme gradient boosting as a classification algorithm. Characteristics using the centralized method produces  features as a central function of protein. Classification results are measured using measurements, precision, recall and ROC scores. Optimizing the model by finding the right parameters produces an accuracy of  and a ROC score of . The prediction model produced by XGBoost has capabilities above the average of other machine learning methods.

Diabetes Mellitus is a serious disease that requires serious treatment. The cause of this disease is due to malfunctions in insulin and insulin-producing organs. One of the proteins that become insulin signaling receptors is IGF1R, which has an important role in activating and maximizing insulin performance. In this study, we aimed to obtain herbal compounds that can activate the function of the IGF1R protein by utilizing compound data in an open database and modeling it using the ensemble method, namely extreme gradient boosting. We found that this method produces the best classification model than with other algorithms. We predicted 844 data for herbal compounds, but only 15 data met the threshold of 0.6. We got one plant from the fifteen herbal compounds, namely Zostera Marine, which was confirmed to have compounds that bind to IGF1R. These compounds have the highest probability value in the classification model that we formed compared to others.

Cancer is a disease that causes an abnormal growth of cells and can attack every part of the body, which is occurred because of a damage in deoxyribonucleic acid (DNA) that leads to a mutation in a vital gene that controls cell division. The biomarker technology that used in clinical practice still used a high cost and need a long time to detect the cancer signs. As the former studies about cancer, the biomarker has been detected in the microarray data. In this paper, we used a support vector machine (SVM) to classify 4 type of leukaemia. Begin with extracting the data feature of sequence DNA from a string into numeric using Second order of Markov chain, SVM classified DNA using 40 data for the training step and 25 data for testing step. In this paper, SVM used 3 types of the kernel, which are linear, Gaussian radial basis function, and polynomial. The results showed that the Gaussian kernel has the best accuracy then other kernel.

Banking is a financial institution tasked with collecting funds from the public and then channeling them back to obtain income. The bank's performance can be seen by comparing the ratio figures in the annual financial statements that the bank has achieved. Therefore, this study aims to determine the CAR, NPL-Net, and LDR, which affect bank profitability or Return on Assets. This research method uses Multiple Linear Regression Analysis with secondary data and a ratio measurement scale, and the number of samples is 90 samples from 30 banking companies during the 2018-2020 period. This data is sourced from the Indonesia Stock Exchange, with the Judgment Sampling technique. The sample was reduced to 84 by the outlier test due to abnormal data. Based on the results of the analysis, it is known that CAR and LDR have no significant effect on Return on Assets, otherwise, NPL-Net has a significant effect on Return on Assets, and CAR, NPL-Net, LDR simultaneously have a significant effect on Return on Assets.

No abstract