Scribd
Upload
519 views15 pages

Time Series Analysis - COMPLETE

This document outlines a summer project on time series analysis and statistical modeling over 8 weeks. The project will cover topics like time series characteristics, components, models including AR, MA, ARMA, GARCH, and state space models. Students will learn to work with time series data in Python, identify patterns, decompose trends and seasonality, fit statistical models and make predictions. The project involves 4 assignments, 2 mini projects involving stock market forecasting and case study analysis, and evaluations at the mid and end terms. The goal is to help students gain practical skills in time series forecasting.

Uploaded by

Lavesh Gupta
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
519 views15 pages

Time Series Analysis - COMPLETE

This document outlines a summer project on time series analysis and statistical modeling over 8 weeks. The project will cover topics like time series characteristics, components, models including AR, MA, ARMA, GARCH, and state space models. Students will learn to work with time series data in Python, identify patterns, decompose trends and seasonality, fit statistical models and make predictions. The project involves 4 assignments, 2 mini projects involving stock market forecasting and case study analysis, and evaluations at the mid and end terms. The goal is to help students gain practical skills in time series forecasting.

Uploaded by

Lavesh Gupta
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 15

STAMATICS SUMMER PROJECTS 2023

TIME SERIES
ANALYIS
& STATISTICAL MODELLING
CONTENT
01 ABOUT TIME SERIES ANALYSIS

02 SYLLABUS

03 PRE-REQUISISTES

04 WEEK WISE PLAN

05 ASSIGNMENTS

06 MID TERM EVALUATION

07 END TERM EVALUATION


ABOUT TSA
In mathematics, a time series is a
series of data points indexed (or
listed or graphed) in time order.

Time series analysis comprises


methods to extract meaningful
statistics and other characteristics
of the data.

Time series forecasting is the use of


a model to predict future values
based on previously observed values.
SYLLABUS
Time Series Data Types of data, Dates and Times, Visualizations, Handling Missing values,
Outlier detection through statistical methods, Resampling data
Smoothening of time series, Decomposition of time series, Stationary
Characteristics processes and detecting/correcting stationarity, Autocorrelation
Internal Structures and components: Definition, Detection and Correction
Components of (i) trend (ii) seasonality (iii) cyclic variations (iv) noise and residuals

Special Models Zero Mean Model, White Noise Model, Random Walk Model, Trend and
Seasonality Models
Univariate and Multivariate AR models: AR, MA, ARMA, ARIMA, Seasonal ARIMA,
AR Models SARIMAX and Vector Models: VAR, VARMA, Confidence Internval

Predicting volatility, Hetereoskedasticity, ARCH Model, GARCH Model,


Volatility Models Special Model: AutoARIMA

Introduction to State Space Models, Multiple Seasonal Trend LOESS


State Space Decomposition for multiple seasonal patterns, Forecasting using
Models Unobverved Components Models (UCM)
PYTHON
ESC111A, ESC112A knowledge but in Python.

PRE BASIC STATISTICS


JEE Level statistics and some adanced topics that can be learnt
REQUISITES quickly, material will be shared.

MTH517A
Just kidding. We'll be covering first half of this
course in our project.
PROJECT TIMELINE
The expected project duration is 8 weeks.
Time commitment: 5-6 hours/week.
Intended audience: Y21, Y22
Pre-reqs: Basic Python and statistics.
4 Assignments + 2 Mini Projects
MTE: 3 Assignments
ETE: 1 Assignment + 2 Mini Projects
Week 1
Time Series Data: Reading Data from CSV, XLXS, URL, APIs, and, Persisting Data
through Pickle, CSV, XLXS

Types of Data: Time Series Data, Cross Section Data, Panel Data

Dates & Time in Python: Working with Pandas DateTimeIndex, Prediction Interval

Visualising Time Series Data: Plotting using Pandas, Interactive visualisation using
hvPlot

Handle Missing Values: Data Quality checks, Univariate Imputation using mean (BFILL,
FFILL, FILLNA), LOCF and NOCB, Multivariate Imputation using Chained Equations,
Linear Interpolation

Outlier Detection: Detection using Visualisation - Box Plots, Inter Quartile Ranges
(IQR), & Kernel Density Estimation (KDE), Detectionn using lag plots,Tukey Method,
Detection using Z-Score, and Kolmogorov-Smirnov test, Detection using QQ Plots
Week 2
Resampling data: Group wise aggregation, Moving Statistics

Windowing Function: Rolling Window, Expanding Window, Exponentially Weighted Moving Window

Lag: Shifting, Lag and Lag Plots


CHARACTERISITICS

Smoothening of time series: First oder, second order smoothing, Simple/Double exponential smoothing,
Triple exponential smoothening

Decomposition: Decomposing time series, LOESS, HP filter, moving averages

Stationarity: Stationary Processes, Differences, Augmented Dickey-Fuller ADF Test, KPSS test, Testing
normality and homoeksdasticity, Detecting and correcting unit roots, Making data stationary (Plots,
Summary statistics, Unit Root Tests)

Autocorrelation: Autorrelation fucntions, ACF and PACF test

Evaluation: Assignment 1
Week 3
Internal Structures of time series: Detecting & Correcting
COMPONENTS

Trend: Definition, Detecting Trend (HP filter), Detrending a time series

Seasonality: Box Plots, Autocorrelation plot, Deseasoning time series,


Seasonal Decomposition

Cyclical Variations: Detecting changes (HP filter)

Residuals: Errors and unexpected variations, Noise

Models for time series: Zero Mean, Random Walk, Trend and Seasonality
Models, White and Red Noise

Evaluation: Assignment 2
Week 4
Time Series Models

Univariate: Auto Regressive (AR), Auto Regressive


Moving Average (ARMA)

Auto Regressive Integrated Moving Average (ARIMA),


Seasonal ARIMA

Multivariate: Vector Autoregression, VARMA,


Confidence Interval

Evaluation: Assignment 3
Week 5
Stochastic Volatility Models

Hetroscedasticity, Predicting Volatility

Models: ARCH (autoregressive conditionally


heteroscedastic) and GARCH

Special Model: Auto ARIMA

Evaluation: Assignment 4
Week 6
Introduction to Space State Models
SPACE STATE MODELS

Decomposing Time series with multiple seasonal


model using Multiple Seasonal Trend decomposition
using LOWESS(MSTL)

Decomposing using Unobserved Component Model


(UCM)
MINI PROJECTS
In Week 7 & 8, we'll be taking up two mini projects to
consolidate all the learnings of the projects. They will
serve as deliverables for evaluation of the project.
At last in our project: an
In this project we'll be
exciting case study on
making predictions on
using Time Series
price of index using
Analysis for
statistical models and
RETROSPECTIVE
we'll be commenting on
ANALYSIS where we'll be
volatility of these markets
focusing on examining
noise/residuals.
STOCK MARKET
FORECASTING CASE STUDY
Week 7 Project 1
Introduction to forecasting, Simple forecasting (returns
with AR and MA)

Intermediate forecasting with MAX models, Advanced


forecasting with Seasonal Models

AutoARIMA forecasting, Pitfalls and Volatility calls

Multiple regression forecasting


Week 8 Project 2

In retrospect approach, examining residuals

Advanced correlation models, correlation plots.

Ensembling of models from scratch, AR between


exogeneous variables

Prediction for the future, Volatility

You might also like