alteryx Use Cases

Learn how you can leverage Alteryx in your organization.
Announcement | Are you an advocate for Alteryx? Check out our Advocacy.Amplified program to get recognized for your evangelism!

Win a Hackathon Without Coding: Predict Bangkok Housing Prices

Community Lead
Community Lead
Created
STelligence_logo.jpg

Overview of Use Case

We are a team of data analysts working with a data consultation firm called STelligence in Bangkok, Thailand. Our company is an operational Intelligence, IT Monitoring, and Big Data Analytics Company. Apart from our day-to-day job, we are always trying to participate in data related contest or hackathon, in order to challenge ourselves. This year we joined a hackathon hosted by HOME dot TECH and Faculty of Engineering Chulalongkorn University called Home Hackathon. The theme of this year’s hackathon is to find the best team that can predict the value of housing in Bangkok given the data they have provided. We used Alteryx to quickly prep the data, alongside its spatial analytic capability to beat the other 15 teams who joined the same contest. This is the 2nd year we won the first place of the hackathon using Alteryx.

 
Describe the business challenge or problem you needed to solve

The challenge we had to solve is to predict and rank the price of 700 properties which are situated in different areas in Bangkok. The data we were given consist of, for example, the size of the land, number of bedrooms and bathrooms, how far from the main road, Lat/Lon, min-max price, distance from POI (shopping mall, train systems etc), the property project name, etc. The objective is to be able to predict the price by ourselves without waiting for the Department of Lands to calculate which normally takes time, or we can also use it as a benchmark. We were given one week to solve the challenge and we knew that with the huge amount of data, we had to spend a lot of time doing data preparation. As we are working professionals, we are left with a few hours after a workday to sort through the data. With Alteryx, we saved valuable time on the data preparation and implementing spatial analysis and we were able to finish the tasks with reliable results.

 
Describe your working solution

There were three data set given to us:

 

  1. Data of 6,000 properties (with prices) to use as a base for modeling
  2. Lat/Lon of POI
  3. Data of 700 properties that we need to predict the price

 

We used Alteryx to prepare the data, for example, using the spatial tool to match the distance of each property to a POI and create a pattern for each distance. From there, we can see how many properties are within each pattern. We were also tasked to research and get more data from other open sources (polygon) to blend them with what we were given (.csv). Moreover, there were some wrong data given to us, for example, some houses are not situated in Bangkok, but in France or in Russia. We used Alteryx to filter out that wrong information before doing further analysis.

 

After we had completed the data preparation and project it into a Tableau dashboard, we used R to do the prediction. While other teams had to spend time coding and using Python, we were able to edge ahead of the competition even without a coding background. with Alteryx’s code-free and code friendly interface we were able to run the predictions much faster – thus winning the hackathon. Alteryx helped us in running the data very quickly. Every time we had to change the logic, we just clicked run and we could continue working with the modeling right away. While other teams who wrote codes used at least 10 minutes to run the data, we ran our data through Alteryx within 1 minute.

 

STelligence1.pngSTelligence2.png

 
Describe the benefits you have achieved

Most of all - Alteryx helped us won the first place of the hackathon for the 2nd year in a row. It’s a big thing for us since we do not have a strong background or experience in writing codes. It made us feel confident that we can do data analysis as good as people who can code. Alteryx made us feel more efficient at work as well. As we need to also serve our clients, we can be confident that we can deliver reliable data to the clients in short time.

 
Contributors
Labels