Predictive Analytics with ARIMA and Prophet in Python

Forecasting time series is a crucial task in business analytics, allowing businesses to anticipate sales, prices, demand, or any variable that evolves over time. In this course, you will learn to apply step-by-step two of the most widely used forecasting models: ARIMA and Prophet.

We begin with an introduction to the fundamentals: what time series are, their components (trend, seasonality, and noise), and how to prepare data by correcting outliers, null values, and duplicates. You’ll see how to use SQL and Python to load and clean real sales data. Then you will dive into modeling with ARIMA, understanding the parameters p, d, q, applying auto_arima, and validating assumptions such as autocorrelation, stationarity, and normality of residuals.

Moving forward, we’ll explore Prophet, a tool developed by Meta that simplifies model creation for predictions. You’ll learn to incorporate exogenous variables (such as promotions and prices), adjust changepoints, customize seasonality parameters and holidays, and generate component graphs to better explain your predictions. We will also compare both models with advanced metrics like AIC and BIC, highlighting the advantages and disadvantages of each approach. The course wraps up with best practices for presenting forecasts in a business context, including confidence intervals, clear visualizations, and storing results in CSV or databases. As a final project, you will automate predictions by exporting models to pickle and scheduling recurring notebooks or scripts. This course combines theory, practice, and real-case scenarios, making it ideal for those looking to master forecasting using ARIMA and Prophet in business settings.

What you will learn:

• Understand time series concepts, trend, seasonality, and noise.
• Detect and correct outliers and common data issues in business.
• Build forecasting models using ARIMA and Prophet from scratch.
• Incorporate exogenous variables such as prices or campaigns to enhance predictions.
• Evaluate models with key metrics: MAPE, RMSE, AIC, and BIC.
• Export and automate predictions in scheduled notebooks or scripts.

Course Content:

• Sections: 8
• Lectures: 39
• Duration: 2h 51m

Requirements:

• Basic knowledge of statistics and time series.
• Familiarity with Python and SQL (basic level).
• Have Python and libraries like statsmodels, pmdarima, and prophet installed.
• Understanding of business analytics (optional but recommended).
• The course is aimed at students and professionals looking to apply forecasting to real business problems.

Who is it for?

• Data analysts, data scientists, and BI professionals.
• Marketing, sales, and finance teams needing reliable projections.
• Students or self-learners interested in machine learning applied to time series.
• Professionals wanting to compare classical methods (ARIMA) with modern approaches (Prophet).

What are you waiting for to get started?

Enroll today and take your skills to the next level. Coupons are limited and may expire at any time!

👉 Don’t miss this coupon! – Cupón DIC25-W2