Companies rely on machine learning engineers to help design and improve the systems that allow their software to improve on its own, rather than being specifically programmed. During the interview process, be prepared to be tested heavily on both computer science and data science knowledge with an emphasis on recognizing patterns and trends. A bachelor's degree in computer science or a related field will be required.
Round 2) Some questions that I remember: Explain the programming assignments, why are the weights randomly initialized, what is an activation function, what is the difference between SoftMax and ReLU activation, what are hyperparameters, list down all the hyperparameters, what are color channels, why do we convert images to grayscale, how does. Otsu's image segmentation work, what is the size of the image before and after converting to grayscale.
Round 3) Some questions that I remember: Explain academic projects, how does an Artificial Neural Network work, what is feed-forward and backward propagation, what is gradient descent, what is the difference between global minimum and local minima, how do you avoid local minima, what is a parameter, difference between parameter and hype parameter, list down each parameter and hyperparameter, mathematical questions on loss function, what is overfitting, how do you avoid overfitting, what is regularization, where do you add regularization term, questions on image classification using CNN, the question to find the second largest element of an array, and a mathematical puzzle.
Can the procedure of tossing a coin be modified to output three outcomes instead of two?
The telephone interview was reasonable was about Machine learning, Estimation theory, software engineering and C++ coding. The on-site interview was a bit biased to software engineering. There was no question about machine learning and it was solely C++11 , design patterns and white board coding. The questions like - name design patterns and show how to use one of them on the board - difference between modern c++ and the classical - what is rule of 5 etc The white board question: - given a text document write a c++ code to retrieve a text
Tell me about your research
assignment carrying question
What would be your ideal style for running a group in this setting?
What's an example of how you dealt with a disagreement with a colleague?
Tell me about yourself. What is your Python experience? Any projects? Why do you want this internship?
SVD (singular value decomposition)
Viewing 1081 - 1090 interview questions