Quiz on Time Series Modeling

Which of the following are true about the Autocorrelation Function (ACF)?

• ACF measures the correlation of a time series with its lagged values.

• ACF can help identify the order of the Moving Average (MA) part in ARIMA models.

• ACF always equals 1 for all lags in a stationary series.

• ACF is used to measure partial correlations in a time series.

What is the purpose of the Partial Autocorrelation Function (PACF)?

• To measure the direct effect of a lag on the current value, excluding intermediate lags.

• To determine the order of the AutoRegressive (AR) part in ARIMA models.

• To identify the trend component in a time series.

• To test if a series is stationary.

Which models are appropriate for forecasting stock price data?

• ARIMA

• Exponential Smoothing

• Linear Regression with trend and seasonality

• LSTM (Long Short-Term Memory neural networks)

When is a Moving Average (MA) model suitable for a time series?

• When the series shows autocorrelation at higher lags.

• When the series exhibits sudden shocks or noise.

• When the ACF shows significant spikes at specific lags.

• When the PACF shows exponential decay.

Which evaluation metrics are commonly used for time series forecasting?

• Mean Absolute Error (MAE)

• Root Mean Square Error (RMSE)

• R-squared (R²)

• Mean Absolute Percentage Error (MAPE)

For analyzing univariate sales data with seasonality and trend, which methods are most suitable?

• Holt-Winters Exponential Smoothing

• SARIMA (Seasonal ARIMA)

• Vector AutoRegression (VAR)

• Random Forest Regression

Which techniques are appropriate for multivariate time series forecasting?

• ARIMA

• VAR (Vector AutoRegression)

• Multivariate Linear Regression

• LSTM

Why might you choose ARIMA over Exponential Smoothing for a specific use case?

• ARIMA models can handle both trend and autocorrelation explicitly.

• ARIMA requires the data to be stationary, while Exponential Smoothing does not.

• ARIMA is easier to implement for non-seasonal data.

• ARIMA is more interpretable for seasonal data patterns.

Which model(s) would you choose for predicting sales data with promotions as an additional variable?

• ARIMAX (ARIMA with exogenous variables)

• Linear Regression

• SARIMA

• LSTM

For evaluating the accuracy of a forecast for a volatile stock price series, which metrics would you prioritize?

• RMSE, as it penalizes larger errors more.

• MAE, as it is more robust to outliers.

• R-squared, as it explains variance in predictions.

• MAPE, as it is useful when the values are non-negative.