Machine Learning Report

Automation and AI

Thanthong Chim-ong Machine Learning Report

Binary Tree 

Data: collect a set of data about the topic that you want your AI to learn. For example, height, age, color.



Convert: after acquire data we need to convert it to something that measurable like 1 for True, 0 for False. Because of AI cannot understand human ideas like color because it is unmeasurable.

Automation and AI



Thanthong Chim-ong

Train: in this process we give the AI a set of converted data and the answer for it to check, after it analyze the data it creates a binary tree. It looks like an upside-down tree. It contains a set of condition (Nodes) and a True, False situations (Lines).

Automation and AI 

Thanthong Chim-ong

Predict: to predict the AI use its binary tree to predict the answer. The tree work by match the information to a condition in the tree if it True it go right and if it False it goes left and go on until it doesn’t have any node to go.

Topic & Data The topic I chose is flowers. The AI has to distinguish between rose, tulip, and Hibiscus uses color, height, and price. Before I change the topic to flowers it was IKEA furniture. I decide to change because the data was too general. Sometime IKEA’s chairs are very similar to table in term of size and price. The AI cannot predict the answer because the actual data that human use to distinguish them are unmeasurable. The challenge was the first and second steps which is collecting data and convert them in to measurable form because I need to find real and accurate information so the AI is real, accurate, and not bias.

Accuracy & Result The AI is surprisingly accurate. It got 100% on the test I gave it. I think because it received enough data to create efficient binary tree. In the measurement sheet tulips are just to distract the AI because I want to know how much information it need to develop an efficient tree. Here are the results.