The project aims to correlate the possibility of Altitude Mountain Sickness (AMS) by analyzing a person's features such as age, smoking habits, consumption of alcohol, and altitude. The project used a paper published in 2008 titled "General Introduction to Altitude Adaptation and Mountain Sickness" by P. Bartsch & B. Saltin to understand the characteristics and causes of AMS. The paper defined the relationships of AMS with altitude, oxygen levels, change in blood hemoglobin levels, and acclimatization concepts. The most crucial element of acclimatization is to maintain efficient O2 delivery in the organs and tissues of the body.
The project used four different classification models, including Naïve Bayes, Decision Tree Classifier, Random Forest Classifier, and KNN Classifier. After analyzing the confusion matrices for each model and iterating through multiple test data, the Random Forest Classifier showed high compatibility with the project's purpose, with an accuracy ranging from 90-95%. The combination model that incorporated the best results from all models was seen to predict the most constant low False Negative and high accuracy.
The project's application can analyze the data of tourists or people visiting high altitude places and predict the possibility of AMS with 92% accuracy. Based on the analysis, the application can provide necessary precautions, medications, and travel advice, such as elevation per day, acclimatization duration, stoppage, type of food, etc., to avoid AMS and have a successful trip.
In conclusion, this reseach used machine learning algorithms to predict the possibility of AMS and developed an application that can help people plan their trip to high altitude places safely.