TIME SERIES ANALYSIS ON
FLIGHT PASSENGERS
Presented By
Harsha
(Reg:2019MCA16090L)
ABSTRACT
● In the real world day–to-day life , Flight is making their own analytics for the
dynamic pricing the fares according to the customer requirement and
decision making on the data .
● Getting insights through data is better than using traditional methods to do
complex calculations and predicting the right one.
● Offering the great discount on sale of fare on non-season market. At peek
time we use some discount with other offers. Maintaining the business
successfully to sustain in Global market .
● Time series analysis is one of the analytics on machine learning. To predict
the future outcomes from past data . To make the works simple by getting
insights from the data. Then we can make things easier. By using statistical
method ARIMA (Auto Regression Integration Moving Average).
CONTENTS
● INTRODUCTION
● EXISTING SYSTEM
● PROPOSED SYSTEM
● SYSTEM REQUIRED SPECIFICATION
● FLOW CHART
● MODEL
● OUTPUT
● CONCLUSION
INTRODUCTION
● Now a days every small entities are tracking their data and doing some
analytics .so the traditional method is difficult to sort high complex problems
so we need to do in different manner which makes our work so important.
● Time Series forecasting is the process of using a statistical model to predict
future values of a time series based on past results.
● It is time dependent. So the basic assumption of a linear regression model
that the observations are independent doesn’t hold in this case.
EXISTING SYSTEM
● Before this analytics came into picture that normal paper based analysis is
there like using some methods and doing all in a vast process and we can’t
do and full of traditional process it will make so many days to predict.
● LIMITATIONS OF EXISTING SYSTEM:
While in Traditional process it will take week’s of time for generating prediction
document.
− More Time
− cost of producing the documents.
− Lack of visualization.
PROPOSED SYSTEM
● Emerging new technologies allow us to do our work on automation and
using some algorithms to predict the future and it will take less time and
paperless .with the help of these tool and packages we can do our work
efficient , fast.
● Like Anaconda distribution we have group of tools to do different kinds of
work for our business analysis and Decision making process .Tools are
Jupyter Notebook, Spyder, Orange etc. We are taking Jupyter Notebook.
From data and applying prediction to the model .
● ADVANTAGES OF PROPOSED SYSTEM
− It doesn’t require manual work.
− Instant output.
SYSTEM REQUIRED SPECIFICATION
SOFTWARE REQUIREMENTS:
− Tool : Anaconda3 Distribution ,Spyder IDLE.
− Language : Python 3.6 or later.
− Packages : pip(19.0 or later)
− Modules : Numpy ,Pandas , Matplotlib , Statsmodels , Datetime.
HARDWARE REQUIREMENT SPECIFICATION:
− Processor : i3 or later versions.
− Speed : 2.0 GHz or more.
− Hard Disk : 256GB
− Monitor : LCD RGB-display .
− Mouse : Logitech.
− RAM : 4.0 Gb.
FLOW CHART
FLOW CHART FOR APPLYING MODEL TO DATA
MODEL
● AR Model: An AR (p) model is an autoregressive model where
specific lagged values of y t are used as predictor variables.
Lags are where results from one time period affect following
periods.
● The value for “p” is called the order.
● It depends on the lagged values of data that you are modelling
to make forecasts.
● AR model is also called conditional models,Markov models or
transition models.
MA Model:The Moving-average model (MA model), also known
as moving-average process, is a common approach for
modeling univariate time series.
It depends on the errors of previous forecasts you made to make
the current forecasts.
The moving-average model specifies that the output variable
depends linearly on the current and various past values.
ARIMA Model: ARIMA, short for ‘Auto Regressive Integrated
Moving Average’.
It is actually a class of models that ‘explains’ a given time series
based on its own past values,that is its own lags and lagged
forecast errors,so that equation can be used to forecast future
values.
An ARIMA model is a class of statistical models for analyzing
and forecasting time series data.
OUTPUT
CONCLUSION
● Now heading in to the Digital World we need to process our data in online .
In the real world day–to-day life .Every Merchant is making their own
analytics for the shop or business according to the customer needs and
decision making on the data .Getting insights through data,Offering the great
discount on sale of goods & services. At peek time we use some discount
with other offers for Maintaining the business successfully to sustain in
Global market .
THANK YOU
