Predicting House Price in Hong Kong #6

It has been a while since I made the last git commit in the project repository. Although I am a bit busy with my daytime job recently, I will keep the side project rolling.

This blog post is about my recent reflection on the project direction.

There are two goals I want to achieve in the project. First, the project can be added to my data project portfolio in which it may help my career. Second, I hope the project can be commercialized. Purchasing property is a big decision. I think the project can add value to Hong Kong people who want to purchase a property.

To achieve the second goal, what is more important is not the valuation mode but whether the product is what users want. In startup world, there is a concept called Minimum Viable Product (MVP). A MVP is a product with sufficient features to satisfy early adopters. It is used to test and validate ideas. I think it is a great idea!.

Minimum Viable Product

Then, what is my MVP?

I think I will use simple machine learning model first, such as K Nearest Neighbor on a limited set of features, such as latitude, longitude, size, floor and building age. Next, I will build a quick prototype of Android App (using Flutter maybe). Finally, I will launch the product to some of my friends and ask them to share. The app will collect basic usage data so that I can analyze the pattern.

To develop such a MVP, it may not take too long. My estimation is 1 to 2 months part time.

That’s what I am going to do in the next few months. Stay tune.