Course Overview
- Course Title: Time Series Analysis & Forecasting
- Instructor: Muhammad Shafiq (Data Scientist, AI & ML Engineer, Lecturer, Researcher)
- Target Audience:
- Aspiring data scientists and analysts
- Business professionals in finance, operations, or marketing
- Students or researchers working with time-dependent data
- Python developers expanding into data forecasting
- Prerequisites:
- Basic Python programming knowledge
- Familiarity with Pandas and NumPy (recommended but not mandatory)
Curriculum Highlights
- Key Topics Covered:
- Fundamentals of time series data (trends, seasonality, cyclical patterns)
- Data preprocessing for time series (handling missing values, smoothing, decomposition)
- Classical forecasting models:
- ARIMA (AutoRegressive Integrated Moving Average)
- SARIMA (Seasonal ARIMA)
- Modern forecasting libraries:
- Facebook Prophet (handling holidays, changepoints, uncertainty intervals)
- Model evaluation metrics (MAE, RMSE, MAPE, AIC, BIC)
- Feature engineering for time series (lag features, rolling statistics)
- Case studies with real-world datasets (e.g., stock prices, sales forecasting)
- Key Skills Learned:
- Implementing time series decomposition (trend, seasonality, residual analysis)
- Building and tuning ARIMA/SARIMA models in Python
- Applying Facebook Prophet for automated forecasting
- Evaluating and comparing forecasting model performance
- Preprocessing and visualizing time-dependent data using Matplotlib/Seaborn
- Deploying forecasting models for business decision-making
Course Format
- Duration:
- 10.5 hours of on-demand video
- 3 practice tests for skill assessment
- Lifetime access to course materials
- Format:
- Self-paced online course (pre-recorded lectures)
- Hands-on coding exercises in Python
- Mobile and TV access for flexible learning
- Resources:
- Downloadable Python notebooks with code templates
- Datasets for practice (CSV/Excel formats)
- Quizzes to reinforce key concepts
- Assignments with real-world forecasting challenges