From Time Series Confusion to Forecasting London’s NO₂: What My First Real Forecasting Project...

2026-07-26 · Source: Data Science on Medium · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics, Environmental Science & Earth Systems · Depth: Intermediate, quick

Summary

This article chronicles a practitioner's initial real-world time series forecasting project, centered on predicting nitrogen dioxide (NO₂) levels in London. The author details the practical challenges encountered, beginning with the crucial task of cleaning imperfect air-quality data. The project involved navigating difficulties with established forecasting methodologies, specifically mentioning struggles with both ARIMA and Prophet models. A significant learning point highlighted is the discovery and application of time-based cross-validation, a technique crucial for robust evaluation in time series contexts. The narrative also encompasses the process of comparing different forecasting approaches to achieve reliable predictions for environmental data.

Key takeaway

For data scientists or AI students embarking on their first real-world time series forecasting project, recognize that initial data quality will likely be imperfect, demanding significant cleaning effort. Do not be discouraged if common models like ARIMA or Prophet present challenges; instead, prioritize discovering and implementing appropriate validation strategies such as time-based cross-validation. This approach will ensure your model comparisons are robust and your predictions for complex data, like environmental metrics, are reliable.

Key insights

Real-world time series forecasting demands robust data cleaning and appropriate validation beyond standard models.

Principles

• Imperfect data necessitates thorough cleaning.
• ARIMA and Prophet models may prove challenging.
• Time-based cross-validation improves robustness.

Method

Clean imperfect data, experiment with models like ARIMA and Prophet, then apply time-based cross-validation for robust comparison.

In practice

• Clean real-world air-quality datasets.
• Test ARIMA and Prophet for NO₂.
• Apply time-based cross-validation.

Topics

• Time Series Forecasting
• Air Quality Data
• NO₂ Prediction
• Data Cleaning
• ARIMA Model
• Prophet Model
• Time-based Cross-Validation

Best for: Data Scientist, AI Student

Related on AIssential

• The Stone in the Lake: What I Learned Trying to Forecast London’s NO2 — Complex time series models do not automatically outperform simpler baselines, and stability across test periods is critical.
• ​​Time Series Cross-Validation: A Guide to Techniques & Practical Implementation — Time series cross-validation maintains chronological order for robust model evaluation and accurate forecasting.
• Modern Forecasting Models: Prophet, Regression/ARIMAX, and Machine Learning — Selecting the optimal forecasting model hinges on data characteristics, complexity, and the need for interpretability or causal explanation.
• WorldReasoner: Evaluating Whether Language Model Agents Forecast Events with Valid Reasoning — WorldReasoner evaluates LM agent forecasting by assessing outcome, evidence, and reasoning quality against temporally valid information.
• Extra #6 - Taking Your RNN From Beginner to Pro — Robust model validation requires more than visual fit, focusing on reproducibility, input visualization, error analysis, and baseline comparison.
• How Far Can You Actually Forecast? The Lyapunov Ceiling — Forecast quality in time series models decays predictably with increasing prediction horizon.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Data Science on Medium.