Data Analytics

Data analytics is a discipline where statistical analysis and technologies being used and applied on dataset to achieve a project objectives. Data analytics often used to find trends or predict the future behaviours based on available dataset so that it can solve the problems stated by a project.

Case Study 1

Predicting Mental Healthiness Among High School Students in Malaysia Using Machine Learning Model

There are three criteria for choosing the best model to be used for this project. Although accuracy is essential, the other two aspects, precision and recall are also being considered as the model deals with the binary classification that will give results whether the student has a mental health-related problem. Recall is the ratio of the number of events you can correctly recall to a number of all correct events, while precision is the ratio of a number of events you can correctly recall to a number of all events you recall (mix of both correct and wrong recalls). Based on the finding, The Random Forest-based model is the most accurate model with 73% accuracy, followed by Decision Tree (72.52%), KNN(66.92%), GBM(65.76%), SVM(64.56%), and Naïve Bayes (56%). The precision for Random Forest is also the best, same percentage with Decision Tree at 73%, followed by KNN(68%), GBM(66%), SVM(65%), and Naïve Bayes (62%). Random Forest-based model also tops the recall percentage, which is important for health-related binary classification model with 73%, the same percentage as the Decision Tree model. The percentages for other models are as follow; KNN(67%), GBM(66%), SVM(65%), and Naïve Bayes (56%). In conclusion, the Random Forest model is the best model for predicting mental health problems amongst high school students in Malaysia.

Case Study 2

Prediction of Quality of Red Wine Using Different Machine Learning Techniques

Based on the table, we can see that Random Forest Classifier has the highest accuracy between the three algotihms tested. This mean that Random Forest Classifier is the best method to be used to predict red wine quality by using machine learning. In recent years, the interest for a good quality red wine keep increasing and the trend is going upwards for upcoming years. The used of machine learning to predict red wine quality can be implemented by big companies.