Amazon Interview Questions

Joshua B. - "reaching my goal and going to the limits"See full answer

Onesimo F. - "N/A. I never had the same experience."See full answer

Shibin P. - "Effective loss functions for computer vision models vary depending on the specific task, some commonly used loss functions for different tasks: Classification Cross-Entropy Loss:Used for multi-class classification tasks. Measures the difference between the predicted probability distribution and the true distribution. Binary Cross-Entropy Loss:Used for binary classification tasks. Evaluates the performance of a model by comparing predicted probabilities to the true binary labe"See full answer

Exponent - "This is an Improve a Product question. Let's first go over the Improve a Product formula: Ask clarifying questions Identify users, behaviors, and pain points State product goal Brainstorm small improvements Brainstorm bolder improvements Measure success Summarize Now, let's begin! Ask clarifying questions Before we begin listing off recommendations, it's important you ask questions to ensure you and the interviewer are on the same page"See full answer

T I. - "This is due to sticky sessions. The load balancer is not correctly configured with sticky session option. It is likely the servers were storing session data on the server themselves (in-memory), and thus when user makes a request, the load balancer routes this to a different server than the one they started with, that second server may not recognise the user's session. This could prompt the user to log in again. One way to resolve this, is to use a centralised session storage, something like"See full answer

Matt M. - "Given the dataset does not contain many labels, it implies we cannot directly use supervised learning. I would ask more about the type of dataset we are given. Is it images, text, etc? This may inform the types of transformations we do the dataset. I can see two approaches to training Given the labels we do have, we can find a method to generate labels for the other unlabeled data. This likely will introduce some error since they may not be true labels, but it at least allows processing the"See full answer

Tony C. - "That I had not been a PM for a formal ML product. But good models need sound data and my five years of sql writing and ensuring that DWHs and marts had the data they needed for reporting would be very relevant"See full answer

Ali H. - "CIDR (Classless Inter-Domain Routing) -- also known as supernetting -- is a method of assigning Internet Protocol (IP) addresses that improves the efficiency of address distribution and replaces the previous system based on Class A, Class B and Class C networks."See full answer

Kegomoditswe R. - "oracle"See full answer