forecasting

Here’s what makes forecasting such a thorny task, and how you can cope with these problemsPhoto by Jukan Tateisi on UnsplashForecasting is a popular but difficult problem in data science.Challenges arise for several reasons, from non-stationarity to noise and missing values. Tackling these issues may be pivotal for improving forecasting performance.Atime series is a sequence of values ordered by time. The key aspect about these data sets is the temporal dependency among observations. What happened in the past affects how…

Create accurate and interpretable predictionsCreated with DALLE According to , Temporal Fusion Transformer outperforms all prominent Deep Learning models for time series forecasting.Including a featured Gradient Boosting Tree model for tabular time series data.But what is Temporal Fusion Transformer (TFT) and why is it so interesting?In this article, we briefly explain the novelties of Temporal Fusion Transformer and build an end-to-end project on Energy Demand Forecasting. Specifically, we will cover:How to prepare our…

A hands-on tutorial on how to apply the Theta model for time series forecasting in PythonPhoto by Hans Reniers on UnsplashWhen it comes to time series forecasting, we often turn our attention to models in the SARIMAX family or exponential smoothing. However, there is one forecasting technique that is rarely mentioned: the Theta model.Despite its simplicity, the Theta model can generate accurate predictions. It performed so well during the M-3 competition, the largest academic time series forecasting competition, that it…

The sky of storm clouds before landfall of Tropical Cyclone Mulan. Credit: Si Gao The South China Sea is where most tropical cyclones (TCs) attack the Chinese mainland, but a lack of observational data has for decades hindered our ability to forecast them. In August 2022, a successful field campaign during TC Mulan boosted confidence in forecasting similar events in the future. The results of the campaign were recorded and…

Studying Events Frequency to Identify Unusual BehaviorsPhoto by Isravel Raj on UnsplashNowadays we live in a data-centric world. With every action, we generate a significant amount of data that can be collected and used to produce valuable business insights. Big tech companies know these dynamics very well. By monitoring our daily activities, it’s possible to identify our habits and preferences to customize offers and increase the probability of engagements.Accessing, and leveraging at the same time, the huge variety of…

Photo: Sarah Silbiger (Getty Images)Why bet on the 2022 World Series when you could bet on the 2022 midterms?That’s the general mindset behind a group of well connected entrepreneurs and moneymakers, including one former Obama White House economist, who are vying to make betting on elections legal in the United States. If successful, the practice would build off surging interest in gambling in recent years, driven in part by the legalization and regulation of online sports betting platforms. At the same time, critics and…

This article shows how to produce multi-step time series forecasts with XGBoost with 24h electricity price forecasting as an example.Photo by Agê Barros on UnsplashA number of blog posts and Kaggle notebooks exist in which XGBoost is applied to time series data. However, it has been my experience that the existing material either apply XGBoost to time series classification or to 1-step ahead forecasting. This article shows how to apply XGBoost to multi-step ahead time series forecasting, i.e. time series forecasting with…

First things first. What is multi-step forecasting?Multi-step forecasting is the problem of predicting multiple values of time series.Figure 1: Forecasts for the next 12 months of total expenditure (billions) on eating out in Australia. Image by Author.Most forecasting problems are framed as one-step ahead prediction. That is, predicting the next value of the series based on recent events. But, forecasting a single step is too narrow for many problems.Predicting many steps in advance has important practical advantages. It…

She and her team flew these drones while they were standing inside the crater to compare faraway atmospheric measurements with those closer to the source. They also used traditional ground-based sampling techniques to collect CO2 directly from the volcano’s gas vents.With their drone data, the researchers found concentrations that were 23 percent higher than usual atmospheric levels, indicating that—despite measuring far from the source—the samples contained enough volcanic CO2 that they could distinguish it in the data.…

This article aims at leveraging time series analysis to predict Power Consumption in the city of Tétouan, MoroccoPhoto by Matthew Henry on UnsplashThe project's goal is to leverage time series analysis to predict energy consumption in 10-minute windows for the city of Tétouan in Morocco.ContextAccording to a 2014 Census, Tétouan is a city in Morocco with a population of 380,000 people, occupying an area of 11,570 km². The city is located in the northern portion of the country and it faces the Mediterranean sea. The…