Review of ML and AutoML Solutions to Forecast Time-Series Data

Area of Science:

• Data Science
• Machine Learning
• Statistical Modeling

Background:

• Time-series forecasting is crucial for predicting future values based on historical data.
• Applications span economics, weather, finance, and business development.
• Traditional linear models and advanced deep learning techniques are employed.

Purpose of the Study:

• To review and analyze various time-series forecasting methodologies.
• To identify challenges and diverse techniques for time-series prediction.
• To serve as a reference for researchers and industries utilizing automated machine learning (AutoML) for forecasting.

Main Methods:

• Comprehensive literature review of time-series analysis techniques.
• Exploration of traditional linear modeling approaches.
• Inclusion of modern automated machine learning (AutoML) and deep learning frameworks.

Main Results:

• A structured overview of the evolution of time-series forecasting methods.
• Identification of current gaps in existing research and techniques.
• Comparison of traditional and advanced forecasting models.

Conclusions:

• AutoML offers promising solutions for complex time-series forecasting problems.
• Further research is needed to address identified gaps in forecasting techniques.
• This review provides a foundation for future advancements in the field.