← Back

Time Series Forecasting

Master the mathematical models behind financial prediction. From stationarity and ARIMA through volatility clustering and GARCH to attention-based transformers.

InputRaw OHLCV
TransformLog Returns
ValidateStationarity Test
ChooseModel Selection
EstimateFit & Forecast
ScoreEvaluate
ShipDeploy

Foundations

Lesson 01

Time Series Basics

Autocorrelation, stationarity, lag plots, and why temporal structure matters for financial data.

Statistics · Math

Lesson 02

Stationarity & Differencing

ADF test, unit roots, differencing, and why models need stationary inputs.

Statistics · Math

Lesson 03

ARIMA

Autoregressive integrated moving average: model order selection, Box-Jenkins, and one-step forecasts.

ARIMA · Python

Volatility

Lesson 04

GARCH

Volatility clustering, ARCH effects, and the GARCH(1,1) conditional variance model.

GARCH · Finance

Lesson 05

Volatility Forecasting

Multi-step variance forecasts, Value at Risk, and asymmetric GARCH variants.

GARCH · Finance

Deep Learning

Lesson 06

Transformers for Time Series

Self-attention, positional encoding, and patch-based sequence modelling.

Deep Learning · Attention

Lesson 07

Full System

Complete forecasting pipeline: ARIMA, GARCH, and Transformer, trained and compared.

Python · Full Build
View the complete forecasting system → arima_model.py · garch_model.py · transformer.py · evaluate.py