[PPT] - Collective Traffic Prediction with Partially Observed Traffic PowerPoint Presentation

SLIDE 1

Motivation Related Works CTP Method Experiments Summary

Collective Traffic Prediction with Partially Observed Traffic History using Location-Based Social Media

Xinyue Liu, Xiangnan Kong, Yanhua Li

Worcester Polytechnic Institute

February 22, 2017

1 / 34

SLIDE 2

Motivation Related Works CTP Method Experiments Summary

About me

2 / 34

SLIDE 3

Motivation Related Works CTP Method Experiments Summary

About me

3 / 34

SLIDE 4

Motivation Related Works CTP Method Experiments Summary

About me

I only know Python (2), and it is great.
I think JavaScript, Ruby, Haskell... are cool, but I am too lazy

to learn them.

I hate C++.

4 / 34

SLIDE 5

Motivation Related Works CTP Method Experiments Summary

My Research Interests

Social Network Analysis [CIKM’16, SDM’17b]
Recommender Systems [SDM’16]
Brain Network [SDM’17a, IJCNN17]

5 / 34

SLIDE 6

Motivation Related Works CTP Method Experiments Summary

Overview

1

Motivation

2

Related Works

3

CTP Method

4

Experiments

5

Summary

6 / 34

SLIDE 7

Motivation Related Works CTP Method Experiments Summary

Why Traffic Prediction?

Excessive traffic causes

travel delays, resource wasting, and pollution.

In 2011, traffic congestion

costs urban Americans 5.5 billion hours of travel delay, 2.9 billion gallons of extra fuel, for a total congestion cost of $121 billion.

7 / 34

SLIDE 8

Motivation Related Works CTP Method Experiments Summary

Why (Location-Based) Social Media?

8 AM 4 PM 11 PM

Temporal Data

Traffic Networks Location-Based Social Media

Traffic Condition

Location Associations

Semantic Data Sensor

“Traffic jam on Storrow Drive, Boston, Massachusetts”

Location-Based Social Media (LBSM) is popular, can be used

as mobile sensors.

Semantic and spatial information from social media can be

helpful.

8 / 34

SLIDE 9

Motivation Related Works CTP Method Experiments Summary

Challenges

Lack of historical traffic data in partial regions.

In real-world road systems, only a small fraction of the road segments are deployed with sensors. It is difficult to predict traffic without traffic history.

Sparsity of LBSM information at fine granularity.

Table: Average # of tweets in each region under different spatiotemporal resolutions Temporal Resolution Spatial Resolution

Ave. #Tweets

12 hours 1 × 1 47,113 1 hour 1 × 1 3,926 1 hour 2 × 2 1,306 1 hour 3 × 3 554 1 hour 4 × 4 389 1 hour 30 × 30 15

9 / 34

SLIDE 10

Motivation Related Works CTP Method Experiments Summary

Conventional Methods

Auto Regression [Smith and Demetsky, 1997, Journal of

Transportation Engineering]

Tweet Semantics [He et al.,2013, IJCAI]

10 / 34

SLIDE 11

Motivation Related Works CTP Method Experiments Summary

Auto Regression [Smith and Demetsky, 1997]

t

time Prediction

spatio-temporal dependencies

Historical Traffic Data

v(t)

g

= α + β1v(t−1)

g

+ β2v(t−2)

g

Fail to work for locations without traffic history.

11 / 34

SLIDE 12

Motivation Related Works CTP Method Experiments Summary

Tweet Semantics [He et al.,2013]

t

time Prediction

Social Media Historical Traffic Data a c e a b c e d a b c e d

Consider each location independently.
Extract tweet semantics as bag-of-words feature for each

location during a 12-hour time window.

Build an auto regression-like model using both traffic history

and tweet semantics.

Fail to work for locations without traffic history.

12 / 34

SLIDE 13

Motivation Related Works CTP Method Experiments Summary

Illustration of CTP [Our Method]

t

time Prediction

congestion spatio-temporal dependencies

time

a b c d e

Local-based Social Media Historical Traffic Data road network a b c e d regions without any sensor

Incorporate LBSM information at finer spatiotemporal

granularity.

Consider different locations collectively.
It works for locations without traffic history!

13 / 34

SLIDE 14

Motivation Related Works CTP Method Experiments Summary

Social Media Semantic Vectors

14 / 34

SLIDE 15

Motivation Related Works CTP Method Experiments Summary

Spatio-temporal Dependencies: I

t-1 t

vi

(t−1)

vj

(t−1)

vq

(t−1)

vp

(t−1)

vi

(t )

vj

(t)

vp

(t)

vq

(t)

Same as the traffic history in auto regression model.

15 / 34

SLIDE 16

Motivation Related Works CTP Method Experiments Summary

Spatio-temporal Dependencies: II

t-1 t

vi

(t−1)

vj

(t−1)

vq

(t−1)

vp

(t−1)

vi

(t )

vj

(t)

vp

(t)

vq

(t)

Spatial dependency within a time window.

16 / 34

SLIDE 17

Motivation Related Works CTP Method Experiments Summary

Spatio-temporal Dependencies: III

t-1 t

vi

(t−1)

vj

(t−1)

vq

(t−1)

vp

(t−1)

vi

(t )

vj

(t)

vp

(t)

vq

(t)

Spatial dependency across time windows.

17 / 34

SLIDE 18

Motivation Related Works CTP Method Experiments Summary

CTP Method

LBSM Semantics t-2 t-1

Training

…

𝑤"

($)

𝑤&

($)

Response 𝑤'

($)

assume time lag = 2 for the simplicity here.
response variable (average speed, total traffic flow, etc).

18 / 34

SLIDE 19

Motivation Related Works CTP Method Experiments Summary

CTP Method

LBSM Semantics Dependency I (Traffic History) t-2 t-1 t-1 t-2

Training

…

𝑤"

($)

𝑤&

($)

Response 𝑤'

($)

19 / 34

SLIDE 20

Motivation Related Works CTP Method Experiments Summary

CTP Method

𝑤"

($%&) 𝑤" ($%()

LBSM Semantics Dependency I (Traffic History) t-2 t-1 t-1 t-2

Training

𝑤)

($%&)𝑤) ($%()

𝑤"

($)

𝑤)

($)

Response 𝑤*

($%&) 𝑤* ($%()

𝑤*

($)

Retrieve the historical data 20 / 34

SLIDE 21

Motivation Related Works CTP Method Experiments Summary

CTP Method

LBSM Semantics Dependency I (Traffic History) Dependency II (Neighbors’ Traffic) t-2 t-1 t-1 t-2 t

Training

…

𝑤"

($)

𝑤&

($)

Response 𝑤'

($)

21 / 34

SLIDE 22

Motivation Related Works CTP Method Experiments Summary

CTP Method

LBSM Semantics Dependency I (Traffic History) Dependency II (Neighbors’ Traffic) Dependency III (Neighbors’ Traffic History) t-2 t-1 t-1 t-2 t t-2 t-1

Training

…

𝑤"

($)

𝑤&

($)

Response 𝑤'

($)

22 / 34

SLIDE 23

Motivation Related Works CTP Method Experiments Summary

CTP Method

LBSM Semantics Dependency I Dependency II Dependency III t-2 t-1 t-1 t-2 t t-2 t-1

Training

Response

Compute using an aggregation function (e.g. average)

Response = Speed, aggregation function = AVG.
𝑤"

($) = 50, 𝑤* ($) = 45, 𝑤, ($) and 𝑤- ($) are unobserved.

The De

Dependency-II II Fea eature for node A at time t is:

(/0

(1) + /2 (1))

3

= 47.5

𝑤*

($)

𝑤6

($)

𝑤"

($)

23 / 34

SLIDE 24

Motivation Related Works CTP Method Experiments Summary

CTP Method

LBSM Semantics Dependency I Dependency II Dependency III t-2 t-1 t-1 t-2 t t-2 t-1

Training

(only observed)

…

Bootstrap

… …

Response (unobserved regions) t-1 t t t-1 t+1 t-1 t

24 / 34

SLIDE 25

Motivation Related Works CTP Method Experiments Summary

CTP Method

LBSM Semantics Dependency I Dependency II Dependency III t-2 t-1 t-1 t-2 t t-2 t-1

Training

(only observed)

…

Bootstrap

… …

Response (unobserved regions) t-1 t t t-1 t+1 t-1 t

25 / 34

SLIDE 26

Motivation Related Works CTP Method Experiments Summary

CTP Method

LBSM Semantics Dependency I Dependency II Dependency III t-2 t-1 t-1 t-2 t t-2 t-1

Training

(only observed)

… … …

Response (unobserved regions) t-1 t t t-1 t+1 t-1 t

Iterative Inference

Keep updating Keep updating

26 / 34

SLIDE 27

Motivation Related Works CTP Method Experiments Summary

Dataset

Traffic Data

Collect from the California Performance Measurement System(PeMS) between October 19 and November 28, 2014. 31,102,272 entries of traffic records.

LBSM Data

Collect tweets from the same area during the same time range using the Twitter streaming API. This collection results in a total number of 2,648,446 tweets.

27 / 34

SLIDE 28

Motivation Related Works CTP Method Experiments Summary

Compared Methods

TDO[Smith and Demetsky, 1997]: Auto regression model using

traffic history.

TDO-floor[——–]: Similar to TDO, except it uses full traffic

history.

TwSeO: A degenerated version of [He et al. 2013], using

tweets semantics.

28 / 34

SLIDE 29

Motivation Related Works CTP Method Experiments Summary

Experimental Setting

Partition the data into two parts, with the beginning (1 − 1

u)

as the training set and the remaining 1

u as the test set

(u = 3, . . . , 7).

k-fold cross-validation is used to randomly sample 1/k regions

as unobserved (k = 2, 3, 4, 5).

Root Mean Square Error (RMSE) is used to evaluate the

performance.

29 / 34

SLIDE 30

Motivation Related Works CTP Method Experiments Summary

Results

low

wer

Is Is be better

ur method
TDO-floor performs the best by using full traffic history.
The proposed CTP outperforms TDO and TwSeO.
The result shows the effectiveness of incorporating tweets

semantics into the collective inference model.

30 / 34

SLIDE 31

Motivation Related Works CTP Method Experiments Summary

The effect of r

low

wer

Is Is be better

ur method

Sparser Information in LBSM Figure: Test Ratio = 1/7 (u = 7)

31 / 34

SLIDE 32

Motivation Related Works CTP Method Experiments Summary

The effect of k

low

wer

Is Is be better

Less Unobserved Regions

ur method

Figure: u = 6, r = 5

32 / 34

SLIDE 33

Motivation Related Works CTP Method Experiments Summary

Summary

Problem Studied

Traffic prediction with partially observed traffic history.

Proposed Model

Using LBSM data to alleviate the issue of absent traffic history. A collective inference model that exploits the complex spatio-temporal dependencies between road segments as well as incorporates LBSM semantics in the prediction.

33 / 34

SLIDE 34

Motivation Related Works CTP Method Experiments Summary

Q&A

Xinyue Liu (xliu4@wpi.edu) Xiangnan Kong (xkong@wpi.edu) Yanhua Li (yli15@wpi.edu)

34 / 34