| Name | Student ID |
|---|---|
| Lam Mai Tuyen | 22521629 |
| Hong Bao Ngoc | 22520953 |
| Nguyeen Thi Thu Uyen | 225211643 |
The objective of this analysis is to explore the cosmetic industry by examining the top three leading brands. The analysis is divided into three main parts, each providing unique insights:
-
Descriptive Analysis: This part focuses on summarizing and visualizing the current state of the cosmetic industry through key metrics and trends observed in the top three brands. It highlights sales performance, market share, and consumer behavior, offering a snapshot of the industry's present landscape.
-
Inferential Analysis: This section uses statistical methods to draw conclusions about the broader industry based on data from the leading brands. It identifies significant patterns, relationships, and differences, providing a deeper understanding of factors driving the industry's performance.
-
Predictive Analysis: The final part employs machine learning and statistical models to forecast future trends in the cosmetic industry. It predicts sales growth, market dynamics, and potential impacts of external factors, offering strategic insights for future planning and decision-making.
This part is performed for all 3 companies: L'Oreal, Unilever and Estee Lauder.
-> Techniques used: Central Tendency, Dispersion Measures, BoxPlot, Candlestick Plot.
This part includes the comparison between 3 companies and find out the top fastest growing companies in terms of their stock price. Then, we use the top to be the main target for further analysis, which are spotting out the existence of correlation between Stock Price and some expernal factors.
-> Techniques used:
- Linear Interpolation for filling in missing data
- MinMaxScaler for scaling the data to the range from 0 to 1
- T- Test (Independent and Paired) for companies comparison
- ANOVA test for correlation test
Model used:
- LSTM Time Series Model
- Decision Tree Regressor
- Random Forest Regressor