Abstract
Data science highlights fields of study and research such as time series, which, although widely explored in the past, gain new perspectives in the context of this discipline. This chapter presents two approaches to time series forecasting, long short-term memory (LSTM), a special kind of recurrent neural network (RNN), and Prophet, an open-source library developed by Facebook for time series forecasting. With a focus on developing forecasting processes by data mining or machine learning experts, LSTM uses gating mechanisms to deal with long-term dependencies, reducing the short-term memory effect inherent to the traditional RNN. On the other hand, Prophet encapsulates statistical and computational complexity to allow broad use of time series forecasting, prioritizing the expert's business knowledge through exploration and experimentation. Both approaches were applied to a retail time series. This case study comprises daily and half-hourly forecasts, and the performance of both methods was measured using the standard metrics.
TopBackground
TS forecasting is an important task that helps organizations with capacity planning. For instance, in the retail business, sales forecasting is behind all strategic and planning decisions. An accurate sales forecast helps with efficient inventory and logistics management.
A TS is a sequence of values measured at successive time instants. The time elapsed between the measurements is usually equally spaced, e.g., minutes, hours, days or months. TS can have three types of components:
Key Terms in this Chapter
Time Series Forecasting: Deals with the estimation of future values using a model that was fit on observations collected over time (time series).
Prophet: An open-source tool provided by Facebook for time series analysis and forecasting. It was designed to facilitate time series forecasting for business experts without deep knowledge in statistics or machine learning.
Machine Learning: A branch of artificial intelligence that studies computational algorithms that adapt their parameters accordingly to a given training dataset.
Recurrent Neural Networks: A class of neural networks that uses feedback loops to model temporal behavior in training data.
Long Short-Term Memory (LSTM): A type of recurrent neural network that uses gating mechanisms to register long-term temporal dependencies in training data.
Retail: Process of selling services or goods using different channels to satisfy consumer needs on identified demands.
Neural Networks: Biological-inspired computational models that mimic the brain structure with simple processing units organized in layers that are highly connected.