Machine learning task You are a data analyst at Amazon and are required to predict if a new dress will be popular (e.g. hitting bestseller list). You can access to a large amount of historical dress selling records at Amazon. Design a machine learning algorithm to solve this problem. Explain which algorithm you plan to adopt and why. Will you choose a supervised or unsupervised method? What kind of data will be required in your method and how are you going to use them? List two attributes or features that could be included in the dataset and explain why you expect them to be useful. List one method which could be used to increase sales Overfitting Explain what overfitting is. How could overfitting be observed? Provide an example how overfitting might happen. Explain how the issue of overfitting could be resolved in terms of building the models, and preparing the data, respectively.
Machine learning task
You are a data analyst at Amazon and are required to predict if a new dress will be
popular (e.g. hitting bestseller list). You can access to a large amount of historical dress
selling records at Amazon.
Design a machine learning
plan to adopt and why.
Will you choose a supervised or unsupervised method?
What kind of data will be required in your method and how are you going to use them?
List two attributes or features that could be included in the dataset and explain why you
expect them to be useful.
List one method which could be used to increase sales
Overfitting
Explain what overfitting is. How could overfitting be observed?
Provide an example how overfitting might happen. Explain how the issue of overfitting
could be resolved in terms of building the models, and preparing the data, respectively.
Step by step
Solved in 2 steps