Difference between revisions of "ANLY482 AY2017-18 T2 Group 31 Main Findings and Analysis"

From Analytics Practicum
Jump to navigation Jump to search
(Created page with "__NOEDITSECTION__ <div align="center"> 500px <!--Header Start--> {|style="background-color:#FFA500; color: #F5F5F5; padding: 10 0 10 0;" width="100%" ce...")
 
 
Line 97: Line 97:
 
<p></p>
 
<p></p>
 
<b>Kernel Density Estimation</b><br>
 
<b>Kernel Density Estimation</b><br>
−
To determine: <br>
 
−
1. To identify cluster of locations that have higher occurrence of indiscriminate parkings<br>
 
  
−
Function (kernel 𝑘) of a given radius (𝑟) “visits” each point in the study region. 𝑘 provides the weight of the area surrounding 𝑠 in proportion to its distance to 𝑠_𝑖 <br>
 
 
<div align="center">
 
<div align="center">
−
[[File:KDEformula.png|200px]]<br>
+
[[File:KEDBedok1.png|500px]]<br>
 
</div>
 
</div>
−
𝑘 is calculated as a function of the distance between point 𝑠 and 𝑠_𝑖, over given radius 𝑟 <br>
+
61 notifications for these clusters<br>
−
The density of the study region is obtained by summing 𝑘 of all points 𝑠_𝑖 within  𝑟 <br>
+
Highest density observed for Blk 761 and 769<br>
 +
Blk 761 next to Bedok Reservoir MRT Station <br>
 +
</div>
 +
 
 +
<div style="margin:20px; padding: 10px; background: #ffffff; text-align:left; font-size: 95%;-webkit-border-radius: 15px;-webkit-box-shadow: 7px 4px 14px rgba(176, 155, 121, 0.96); -moz-box-shadow: 7px 4px 14px rgba(176, 155, 121, 0.96);box-shadow: 7px 4px 14px rgba(176, 155, 121, 0.96);">
 +
<p></p>
 +
<b>Observational Study</b><br>
 
<div align="center">
 
<div align="center">
−
[[File:LargeBW.png|400px]] [[File:SmallBW.png|400px]]
+
[[File:Bedok1.png|500px]]<br>
 
</div>
 
</div>
−
Kernel Density Estimations are sensitive to changes in radius values <br>
+
Next to Bedok Reservoir MRT Station <br>
−
Large radius leads to a smoother curve, but local details would be obscured <br>
+
Long cycling path along the perimeter of the HDB precinct to the MRT Station<br>
−
Small radius leads to many small spikes that are very localised <br>
+
Convenient to cycle from their home to the MRT Station<br>
−
Using the statistically significant radius distance obtained from Modified L Test as a search radius within each event<br>
+
Park the bikes at Block 761 before going MRT<br>
  
 
<div align="center">
 
<div align="center">
−
[[File:Interpolate.png|400px]]
+
[[File:Bedok2.png|500px]]<br>
 +
</div>
 +
Connected to Carpark Deck 3A via linkway bridge<br>
 +
Overhead bridge that connects from Deck 3A to Fengshan precincts<br>
 +
Through interviews, people use the overhead bridge to get home from Fengshan Market & Food Centre<br>
 +
 
 +
<b>Effectiveness of Search Radius</b>
 +
<div align="center">
 +
[[File:Bedok3.png|500px]]<br>
 +
</div>
 +
Within 62m search radius, able to find an indiscriminately parked bike most of the time<br>
 +
However, need to consider cases of staircase and slopes<br>
 +
Inconvenience of carrying bikes up<br>
 +
Should still check these spots even if the distance is further than 62 metres<br>
 +
 
 +
<b>Lack of Yellow Boxes</b>
 +
<div align="center">
 +
[[File:Yellow.png|500px]]<br>
 +
</div>
 +
Only 2 yellow boxes at Blk 765 and Blk 770<br>
 +
Yellow boxes are 220 metres away from each other but entire precinct is 630 metres<br>
 +
Resident mentioned that the problem was worse in the past and more yellow boxes would help to alleviate the problem<br>
 
</div>
 
</div>
−
Perform interpolation by transforming the graph to make it smoother<br>
 
−
Individual kernels are summed up to produce a smooth surface<br>
 
−
Quartic kernel type is used in QGIS<br>
 

Latest revision as of 22:56, 14 April 2018

Bannernew.png

HOME

 

ABOUT US

 

PROJECT OVERVIEW

 

PROJECT ANALYSIS

 

PROJECT MANAGEMENT

 

ANLY482 HOMEPAGE

Data

Exploratory Data Analysis

Model Building

Main Findings and Analysis

Recommendation

Findings and Analysis

Spatial Point Analysis

SPA.png

Conducted spatial point distributions in Singapore on QGIS
Focus on HDB Land-Use Type as highest number of notifications
Case study done on Bedok Reservoir HDB region

Bedok.png

Belongs to the same precinct with only one entrance at Bedok Reservoir View near roundabout

Modified L Test

ModLTest1.png

Sharp increase due to data quality issues during collection of data
Duplicates and points that are very close together
Signs of statistically significant clustering even at very small radius

ModLTest2.png

Between 54.5 metres and 62 metres
Clustering but not statistically significant

ModLTest3.png

Statistically significant radius of 62 metres with signs of clustering
Likely to find another indiscriminately parked bike within 62 metres
Used as input for kernel density estimation

Kernel Density Estimation

KEDBedok1.png

61 notifications for these clusters
Highest density observed for Blk 761 and 769
Blk 761 next to Bedok Reservoir MRT Station

Observational Study

Bedok1.png

Next to Bedok Reservoir MRT Station
Long cycling path along the perimeter of the HDB precinct to the MRT Station
Convenient to cycle from their home to the MRT Station
Park the bikes at Block 761 before going MRT

Bedok2.png

Connected to Carpark Deck 3A via linkway bridge
Overhead bridge that connects from Deck 3A to Fengshan precincts
Through interviews, people use the overhead bridge to get home from Fengshan Market & Food Centre

Effectiveness of Search Radius

Bedok3.png

Within 62m search radius, able to find an indiscriminately parked bike most of the time
However, need to consider cases of staircase and slopes
Inconvenience of carrying bikes up
Should still check these spots even if the distance is further than 62 metres

Lack of Yellow Boxes

Yellow.png

Only 2 yellow boxes at Blk 765 and Blk 770
Yellow boxes are 220 metres away from each other but entire precinct is 630 metres
Resident mentioned that the problem was worse in the past and more yellow boxes would help to alleviate the problem

Retrieved from ‘https://wiki.smu.edu.sg/ANLY482/index.php?title=ANLY482_AY2017-18_T2_Group_31_Main_Findings_and_Analysis&oldid=23671’