2025_T2_ML_Team(ARIMA Model)

[Denissss1213]

Data Synthesis Completion and Exploratory Data Analysis (EDA):

1. Successfully generated comprehensive synthetic price data for Coles supermarkets spanning 2024-2025, creating a robust dataset of 22,375 unique products across 13 main categories

2. Combining the price multiplier for a specific month and the pattern for a specific category, adjusts the price changes for seasonal categories to within ±20%.

3. Conducted comprehensive univariate analysis revealing price distribution characteristics: mean=$10.95, with a right-skewed distribution indicating premium products

4. Discovered strong category-level differentiation: Meat & Seafood averaged $18.20 while Pantry items averaged $6.45

5. Correlation analysis revealed expected inverse relationship between price and discount rate (r=-0.43)

ARIMA Modeling Research Exploration:

1. Conducted initial exploratory analysis to understand time series characteristics of supermarket pricing data
   2 .Perform comprehensive stationarity tests to confirm that the data are suitable for ARIMA modeling
   3 .Analyzing potential autocorrelations (AR) and moving averages (MAs) reveals clear seasonal patterns in pricing data

ARIMA Model Step:

1. Data Preparation
2. Stationarity Test
3. ACF/PACF Analysis 4.Training/testing set split
4. Initial Model – ARIMA(2,1,0)
5. Model Comparison and Optimization 7.Future Forecast (30 days)