Data Science and Machine Learning (Theory and Projects) A to Z - RNN Architecture: Fixed Length Memory Model Exercise So

The video tutorial explains the concept of averages, focusing on weighted averages with parameters alpha and beta. It discusses sequence modeling in stock price prediction, emphasizing the use of running averages to incorporate historical data. The importance of hyperparameters in determining the contribution of past and current data is highlighted. The tutorial contrasts classical methods like moving averages with modern approaches such as recurrent neural networks for stock market predictions.

What is the role of alpha in the weighted average formula?

• It represents the contribution of historical data.

• It represents the contribution of the current data point.

• It is used to normalize the average.

• It is a constant value used for all calculations.

In sequence modeling, why might we look at data points from several steps back?

• To reduce the influence of recent data.

• To capture trends from recent history for better predictions.

• To ensure all data points are equally weighted.

• To increase the complexity of the model.

How does a running average help in stock price prediction?

• By averaging only the most recent two data points.

• By using only the first and last data points in the sequence.

• By incorporating all past data points to provide a comprehensive prediction.

• By ignoring historical data and focusing only on the latest data point.

What is a key consideration when deciding the values of alpha and beta in running averages?

• Adjusting them based on the importance of recent versus historical data.

• Maximizing the contribution of historical data.

• Ensuring both are equal to maintain balance.

• Keeping them constant for all types of data.

Which method is mentioned as a classical approach to stock price prediction?

• Decision trees.

• Moving averages.

• Linear regression.

• Recurrent neural networks.