Exam AWS Certified Machine Learning - Specialty topic 1 question 160 discussion

Answer is C, CNN-QR and DeepAR accepts related time series data (weather data, number of people on property, etc.,)

Option C: Convolutional Neural Network - Quantile Regression (CNN-QR). This algorithm is well-suited for handling complex datasets with multiple features, such as historical power consumption, weather, number of individuals, and public holidays, providing accurate and robust forecasts.

Answer is C, CNN-QR and DeepAR accepts related time series data (weather data, number of people on property, etc.,). Classic forecasting methods, such as ARIMA or exponential smoothing (ETS), fit a single model to each individual time series. In contrast, DeepAR+ creates a global model (one model for all the time series) with the potential benefit of learning across time series. Source: https://aws.amazon.com/blogs/machine-learning/making-accurate-energy-consumption-predictions-with-amazon-forecast/

As per https://docs.aws.amazon.com/forecast/latest/dg/aws-forecast-choosing-recipes.html A. NO - no as powerful as NN B. NO - no as powerful as NN C. NO - works best with 100's of time series D. YES - best for strong seasonnability, expected for power

Based on this only CNN-QR can accept historical data https://www.examtopics.com/exams/amazon/aws-certified-machine-learning-specialty/view/32/

CNN-QR is a deep learning algorithm that can model complex relationships between the inputs and outputs, such as the weather and public holidays, with historical power consumption data. CNN-QR has been shown to be effective in generating accurate predictions in many different types of forecasting use cases, including demand forecasting. ETS (Exponential Smoothing) is a classical time series algorithm that is often used for forecasting. It can be effective for simple time series data that have regular patterns, but may not be sufficient to handle the complexity of the given data. ARIMA (Autoregressive Integrated Moving Average) is another classical time series algorithm that can model complex patterns in data. However, it may be difficult to use in cases where there are many different inputs and the relationships between the inputs and outputs are complex.

ARIMA & ES are both base time series algos that are available. DeeoAR+ & CNN-QR are refined and able to utilize external data as well to complement the time series data available

C, as explained here: https://docs.aws.amazon.com/forecast/latest/dg/aws-forecast-choosing-recipes.html

According to the link below, it is either ARIMA or DeepAR. So A is the answer here https://aws.amazon.com/blogs/machine-learning/making-accurate-energy-consumption-predictions-with-amazon-forecast/

Given the provided data, I would discard A and B. Amazon Forecast CNN-QR, Convolutional Neural Network - Quantile Regression, is a proprietary machine learning algorithm for forecasting scalar (one-dimensional) time series I would choose D, Prophet. https://docs.aws.amazon.com/forecast/latest/dg/aws-forecast-recipe-prophet.html How Prophet Works Prophet is especially useful for datasets that: Contain an extended time period (months or years) of detailed historical observations (hourly, daily, or weekly) Have multiple strong seasonalities Include previously known important, but irregular, events Have missing data points or large outliers Have non-linear growth trends that are approaching a limit Prophet is an additive regression model with a piecewise linear or logistic growth curve trend. It includes a yearly seasonal component modeled using Fourier series and a weekly seasonal component modeled using dummy variables.