Introduction
Our group consists of 3 team members:
Low Li Ci 206476
Long Yeh Min 206595
Aymane Igmiden 204807
This project used the elderly dataset that is provided by the first company using various instruments for mild cognitive impairment. The data is collected from 4 states, which are Kelantan, Perak, Selangor and Johor, and categorised into 4 domains: Demographic, Social, Health and Psychology.
Objectives
Predict the score of MMSE, which is a clinical examination to measure cognitive impairment
Predict the score GDS-15, which is used to screen, diagnose, and evaluate depression in elderly individuals.
Create a visualisation dashboard to analyse important features that affect MMSE and GSD-15
Produce a data product of our machine learning model by using streamlit
Stakeholders
Company:
Able to know the factors that affect MMSE and GDS-15 scores the most
Able to visualise the factors in different domains by using our dashboard
Able to predict the mental state and geriatric depression of elderly by using our data product
Hospital:
Able to conduct preliminary prediction of the MMSE and GDS-15 scores of elderly patients by using our data product
Able to know the trend of the people getting severe MMSE and GDS-15 by using our dashboard.
Questions
Which factors affect MMSE and GDS-15 scores the most?
Which domain has the biggest impact on MMSE and GDS-15?
Which state has the highest score for MMSE?
Which gender has the highest score for MMSE?
Do smokers or alcoholics likely to have a higher score for MMSE?
Does MMSE and GDS-15 affect each other?