The dataset contains data obtained from students regarding their weekly usage of social media. Subsequently, the data has been consolidated into a single row through averaging. It includes multiple other target features such as mood productivity, whether or not a student had trouble falling asleep, tired waking up in the morning, and the energy level throughout the week.
Every week students were expected to log the total hours spent on different applications, and fill out details related to whether they were productive, or had trouble with their sleeping pattern, their mood throughout the week, and more.
| Applications Tracked |
|---|
| WhatsApp/WeChat |
| Snapchat |
| YouTube |
| OTT Platforms |
The approach employed for this analysis involves the following steps:
Principal Component Analysis- Used to explore the potential reduction of dimensions by determining the optimal number of Principal Components (PCs).Clustering- Utilized to identify patterns within the data by grouping similar data points.Factor Analysis- Employed to identify underlying factors that can group various columns.Multiple Regression and Logisitic Regression- Resorted to Regression to predict whether students' face trouble sleeping based on social media usage.Linear Discriminant Analysis- Applied LDA to create a model to predict whether students' face trouble sleeping or not and the correlation of the same to their social media usage
The cleaned data, Rmd (R), and HTML files are available to play with!