6680172_​DALDailyStockPrice2025

Data Collection 

- Load DAL daily stock price data Jan 2 - Jun 30, 2025 no Jan 1 from Investing.com

- Convert Date column from String to Date and Time

- Sort data ascending by date (oldest to newest)

Data Cleaning & Preparation 

- Remove unnecessary columns, only keeping Date and Price

- Apply log transformation

Outlier Detection 

- Calculate 21-day rolling mean and standard deviation 

- Compute UCL and LCL

- Flag any values outside UCL/LCL as outliers

Holt's Double Exponential Smoothing

- Fit Level + Trend model on Log_Price

- Optimize alpha and beta using brute force search

- Evaluate using RMSE via Numeric Scorer

Holt's Triple Exponential Smoothing 

- Fit Level + Trend + Seasonality model on Log_Price 

- Seasonal period m=5 (trading week) 

- Optimize alpha, beta and gamma using brute force search

 - Evaluate using RMSE via Numeric Scorer

ARIMA Modeling 

- Split data 80/20

- Search best p, d, q using brute force grid search 

- Train ARIMA Learner on 80% training set 

- Predict on 20% test set using ARIMA Predictor 

- Evaluate using RMSE via Numeric Scorer

Exploratory Visualization

 - Compare the effect of log transformation with the Raw Price

Model Comparison

- Compare RMSE of all three models:

Holt's Double = 0.059

Holt's Triple = 0.053

ARIMA(0,0,0) = 0.062

- Holt's Triple performed best which has the lowest error

Conclusion

Holt's Triple Exponential Smoothing performed best with the lowest RMSE of 0.053, followed by Holt's Double (0.059) and ARIMA (0.062).

Holt's Triple outperformed the others because it captures the weekly seasonal pattern (every 5 days) in DAL's daily trading data.

ARIMA(0,0,0) performed worst, suggesting DAL's stock prices is difficult to predict using past values alone.

Therefore, Holt's Triple is recommended for forecasting DAL's daily stock price.