Validated method for automated glioma diagnosis from GFAP immunohistological images: a complete pipeline

Background and objective: Glial fibrillar acid protein is a common marker for brain tumor because of its particular rearrangement during tumor development. It is commonly used in manually histological glioma detection and grading. An automatic pipeline for tumor diagnosis based on GFAP is proposed in the present manuscript for detecting and grading canine brain glioma in stages III and IV. Methods: The study was performed on canine brain tumor stages III and IV as well as healthy tissue immunohistochemically stained for gliofibrillar astroglial protein. Four stereological indexes were developed using the area of the image as reference unit: density of glioma protein, density of neuropil, density of astrocytes and the glioma nuclei number density. Images of the slides were subset for image analysis (n=1415) and indexed. The stereological indexes of each subset constituted an array of data describing the tumor phase of the subset. A 5% of these arrays were used as training set for decision tree classification with PCA. The other arrays were further classified in a supervised approach. ANOVA and PCA analysis were applied to the indexes. Results: The final pipeline is able to detect brain tumor and to grade it automatically. Added to it, the role the neuropil during tumor development has been quantified for the first time. While astroglial cells tend to disappear, glioma cells invade all the tumor area almost to a saturation in stage III before reducing the density in stage IV. The density of the neuropil is reduced during the tumour growth. Conclusions: The method validated ere allows the automated diagnosis and grading of glioma in dogs. This method opens the research of the role of the neuropil in tumor development.

Motivation, purpose, and purchasing frequency of honey consumption in West Java

Indonesia is home to the greatest diversity of social bees in all over Asia, particularly species of the genus Apis. Thus, expanding the apiculture industry for commercial development is highly considerable. Although this industry has not become a special concern, the products of this industry are very popular among the Indonesian people, both for health, lifestyle, and other benefits. Research plays an essential role for good decision making, however, there is little research related to honey marketing in Indonesia. In this study, we observed the honey consumption of 246 respondents living in West Java by using online questionnaires and Decision Tree Classification to contribute to honey marketing research. This research shows that the motivation of the respondents in buying honey was merely for health reasons and the main purpose was for personal consumption. As for purchasing frequency, 86% of respondents purchased honey more than once a month. Then, a classification model of honey purchasing frequency based on respondents’ demographics which has an accuracy of 70.3% was built. The study results should be considered by the food industry and honey producers to emphasize consumer behaviour to formulate a better marketing strategy.

A COMPARATIVE STUDY OF THE DIFFERENT CLASSIFICATION ALGORITHMS ON FOOTBALL ANALYTICS

Sports analytics is on the rise, with many teams looking to use data science and machine learning algorithms to augment their teams research and boost team performance. This is especially true in the case of Football Clubs. In this work, we have taken the statistics of matches for each team from five major football leagues. These include the English Premier League, La Liga, Serie A, Bundesliga, and Ligue 1. We use this data for two kinds of classification to predict a teams win, loss, or draw. First, we implement Multiclass Classification using Naive Bayes classification, Decision Tree classification, and K-Nearest Neighbours classification. We use f1-score, recall, and precision to evaluate the model. Next, we use Binary Classification to predict if a team wins or does not win, i.e., a loss or a draw. We achieve this by using Support Vector Machines, Logistics Regression, K-Nearest Neighbours classification, Decision Tree classification, and Naive Bayes classification. We evaluate the results using the evaluation metrics mentioned above. Now, we compare the accuracy and efficacy of these algorithms based on the evaluation metrics. This will help standardize the means of classification in sports and football analytics in the future.

Decision tree classification of digital soil, weather, crop mapping and yield prediction using linear regression with region influences.

This project deals with the study of soil properties, crop and the regional influences along with their dependencies which would be further used for a digital map. Both classification and regression algorithms were carried out and a decision tree as well as a decision regressor tree was plotted to finalise the results. Out of the 6 classification algorithms applied decision tree gave the highest accuracy of 95.24% and linear regression gave the best accurate results of 100% among the 3 regression algorithms.