Scribd
Upload
10 views48 pages

MSCA 691 Lecture 2

The document outlines the content of MSCA 692: Advanced Data Mining, focusing on time series forecasting techniques such as multiple linear regression, polynomial regression, and trend-based models. It discusses methods for estimating coefficients, including regression with lagged values and various curve fitting approaches. Additionally, it covers autoregressive and moving average models, as well as ARIMA model identification.

Uploaded by

heliajavanmard2000
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
10 views48 pages

MSCA 691 Lecture 2

The document outlines the content of MSCA 692: Advanced Data Mining, focusing on time series forecasting techniques such as multiple linear regression, polynomial regression, and trend-based models. It discusses methods for estimating coefficients, including regression with lagged values and various curve fitting approaches. Additionally, it covers autoregressive and moving average models, as well as ARIMA model identification.

Uploaded by

heliajavanmard2000
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

MSCA 692: Advanced

Data Mining
John Molson School of Business,
Concordia University
Teacher: Salim Lahmiri

MSCA 691 - Lecture 2 1


MSCA 691 - Lecture 2 2
Time series forecasting

1)Multiple linear
regression
2) Polynomial Regression
3) Trend-based models

MSCA 691 - Lecture 2 3


MSCA 691 - Lecture 2 4
Patterns to
look for

MSCA 691 - Lecture 2 5


MSCA 691 - Lecture 2 6
MSCA 691 - Lecture 2 7
Examples of
performance
measures

MSCA 691 - Lecture 2 8


MSCA 691 - Lecture 2 9
MSCA 691 - Lecture 2 10
MSCA 691 - Lecture 2 11
MSCA 691 - Lecture 2 12
MSCA 691 - Lecture 2 13
MSCA 691 - Lecture 2 14
MSCA 691 - Lecture 2 15
MSCA 691 - Lecture 2 16
MSCA 691 - Lecture 2 17
MSCA 691 - Lecture 2 18
MSCA 691 - Lecture 2 19
MSCA 691 - Lecture 2 20
MSCA 691 - Lecture 2 21
MSCA 691 - Lecture 2 22
MSCA 691 - Lecture 2 23
MSCA 691 - Lecture 2 24
MSCA 691 - Lecture 2 25
MSCA 691 - Lecture 2 26
MSCA 691 - Lecture 2 27
MSCA 691 - Lecture 2 28
MSCA 691 - Lecture 2 29
Estimation of coefficients
Model to be estimated

Error of the model

Cost function to minimize

Calculating
derivatives

MSCA 691 - Lecture 2 30


Estimation of coefficients

MSCA 691 - Lecture 2 31


Regression with lagged values of input X

1)Multiple regression model with current


and past values (lags) of X used as
additional inputs.

2)q = lag length = lag order

3)OLS algorithm is used for estimation


MSCA 691 - Lecture 2 32
Regression with lagged values of output Y

Sometimes a past
value of Y is used
as a predictor as
well.

MSCA 691 - Lecture 2 33


How to fit
theses curves ?

MSCA 691 - Lecture 2 34


Curve fitting
Basic model

Polynomial
regression

Reciprocal fit

MSCA 691 - Lecture 2 35


Curve fitting
Basic model

Quadratic
trend

S-curve trend

MSCA 691 - Lecture 2 36


Curve fitting
Basic model

Exponential
growth

MSCA 691 - Lecture 2 37


Potential fitting
and predictive
models

MSCA 691 - Lecture 2 38


Potential fitting
and predictive
model

MSCA 691 - Lecture 2 39


Potential fitting
and predictive
model

MSCA 691 - Lecture 2 40


Potential fitting
and predictive
model

MSCA 691 - Lecture 2 41


Autoregressive
model

MSCA 691 - Lecture 2 42


Autoregressive
model

MSCA 691 - Lecture 2 43


Moving
average model

MSCA 691 - Lecture 2 44


Moving
average model

MSCA 691 - Lecture 2 45


ARIMA
model

MSCA 691 - Lecture 2 46


Model
identification

MSCA 691 - Lecture 2 47


Model
identification

MSCA 691 - Lecture 2 48

You might also like