This experiment tries to ingest a Dota2 match data with heroes and creates a predictive webservice to predict the odds of winning. It was created as the very basic and simple demonstration of how to create projects in Azure ML studio.
This experiment tries to ingest a Dota2 match data with heroes and creates a predictive webservice to predict the odds of winning vs hero composition. This does not aim to demonstrate maximum accuracy or any other complex ML concepts such as the use of algorithms or tuning hyperparameters. The train module used was the simplest as to make the live demo quick and easy to understand. I got the data from: https://archive.ics.uci.edu/ml/datasets/Dota2+Games+Results I then asked the question of: **What team composition would i recommend in playing ranked games against South Korea region?** Engineering the data, I referenced the feature labels based from the code below: *https://gist.github.com/da-steve101/1a7ae319448db431715bd75391a66e1b* -and the Dota2 API for the other information: *http://dota2api.readthedocs.io/en/latest/responses.html* The data rows consisted of: Each row of the dataset is a single game with the following features (in the order in the vector): 1. Team won the game (1 or -1) 2. Cluster ID (related to location) 3. Game mode (eg All Pick) 4. Game type (eg. Ranked) 5 - end: Each element is an indicator for a hero. Value of 1 indicates that a player from team '1' played as that hero and '-1' for the other team. Hero can be selected by only one player each game. This means that each row has five '1' and five '-1' values. In the sketch, I also kind of queried the data specific with South Korea region and Ranked games.