Model Selection Suggestions
Selecting Models for Time-Series Forecasting
This prompt helps data science teams choose the most appropriate models for time-series forecasting. It focuses on understanding trends, seasonality, and other temporal features to select models that deliver accurate predictions while accommodating the dataset’s constraints.
Responsible:
Data Science
Accountable, Informed or Consulted:
Data Science, Engineering
THE PREP
Creating effective prompts involves tailoring them with detailed, relevant information and uploading documents that provide the best context. Prompts act as a framework to guide the response, but specificity and customization ensure the most accurate and helpful results. Use these prep tips to get the most out of this prompt:
Analyze the dataset for trends, seasonality, and patterns using visualization techniques.
Define the forecasting horizon and frequency (e.g., daily, monthly).
Identify any domain-specific requirements, such as sensitivity to holidays or events.
THE PROMPT
Help recommend suitable models for time-series forecasting using [specific dataset, e.g., sales data with seasonal trends]. Focus on:
Basic Models: Suggesting foundational approaches, such as, ‘For simple trends, recommend models like ARIMA or Holt-Winters Exponential Smoothing.’
Seasonal and Non-Linear Trends: Recommending advanced models, such as, ‘For datasets with seasonality or non-linear patterns, suggest Seasonal ARIMA (SARIMA), Prophet, or LSTMs (Long Short-Term Memory Networks).’
Multivariate Forecasting: Including complex datasets, such as, ‘For datasets with multiple influencing variables, propose models like Vector Autoregression (VAR) or NeuralProphet.’
Real-Time Applications: Proposing scalable solutions, such as, ‘For real-time forecasting needs, recommend lightweight models like exponential smoothing or frameworks like AWS Forecast.’
Validation Strategy: Suggesting evaluation techniques, such as, ‘Use rolling window validation or time-series cross-validation to ensure robust performance across different time periods.’
Provide tailored model recommendations that align with the time-series forecasting requirements and constraints. If additional details about trends, seasonality, or evaluation metrics are needed, ask clarifying questions to refine the suggestions."
Bonus Add-On Prompts
Propose methods for preprocessing data with missing timestamps or irregular intervals before applying forecasting models.
Highlight techniques for handling outliers and anomalies in time-series data.
Suggest strategies for combining statistical and machine learning models for hybrid forecasting approaches.
Use AI responsibly by verifying its outputs, as it may occasionally generate inaccurate or incomplete information. Treat AI as a tool to support your decision-making, ensuring human oversight and professional judgment for critical or sensitive use cases.
SUGGESTIONS TO IMPROVE
Focus on models for specific time-series types, such as financial markets or sensor data.
Include tips for feature engineering, like lagged variables or rolling statistics.
Propose ways to integrate exogenous variables for improving forecast accuracy.
Highlight tools like statsmodels, TensorFlow, or PyCaret for time-series modeling.
Add suggestions for automating hyperparameter tuning for models like ARIMA or Prophet.
WHEN TO USE
During the initial selection phase for time-series forecasting projects.
To compare different forecasting approaches for datasets with varying complexity.
When refining models for applications with temporal dependencies.
WHEN NOT TO USE
For static datasets without temporal information.
If trends and seasonality are minimal or non-existent in the data.