i
Chubb
Filter interviews by
I applied via Naukri.com and was interviewed in May 2022. There were 4 interview rounds.
Recall is a metric used to measure the ability of a model to find all relevant instances in a dataset.
Recall is the ratio of true positives to the sum of true positives and false negatives.
It is used as an evaluation metric to assess the model's ability to identify all positive instances correctly.
In multi-label data, recall can be calculated for each label separately and then averaged.
The number of hidden layers and n...
Universal approximation theorem states that a neural network with a single hidden layer can approximate any continuous function.
A neural network with a single hidden layer can approximate any continuous function
It is a fundamental theorem in the field of deep learning
The theorem applies to a wide range of activation functions
The number of neurons required in the hidden layer may vary depending on the complexity of the ...
Backpropagation is a supervised learning algorithm used to train neural networks by adjusting weights to minimize error.
It involves propagating the error backwards through the network to adjust the weights of the connections between neurons.
The algorithm uses the chain rule of calculus to calculate the gradient of the error with respect to each weight.
The weights are then updated using a learning rate and the calculate...
Avoid ReLU when dealing with negative values or vanishing gradients.
When dealing with negative values, use Leaky ReLU or ELU instead.
When facing vanishing gradients, use other activation functions like tanh or sigmoid.
In some cases, using ReLU in all layers can lead to dead neurons.
Consider the nature of your data and the problem you are trying to solve before choosing an activation function.
To get embeddings of long sentences/paragraphs truncated by BERT, we can use pooling techniques like mean/max pooling.
We can use pooling techniques like mean/max pooling to get embeddings of truncated sentences/paragraphs.
We can also use sliding window approach to get embeddings of overlapping segments of the long input.
For using BERT on such long inputs, we can use sentence embeddings or word embeddings depending on t...
I would want to be a dolphin because of their intelligence, social nature, and ability to swim freely in the ocean.
Dolphins are known for their high level of intelligence and problem-solving abilities.
They are highly social animals and live in groups called pods.
Dolphins have the ability to communicate with each other using a complex system of clicks, whistles, and body movements.
They are also known for their agility a...
Lack of communication, unresponsiveness, and lack of accountability frustrate me.
Lack of communication: When teammates fail to communicate important information or updates, it hampers the progress of the project.
Unresponsiveness: When teammates do not respond to emails, messages, or requests in a timely manner, it slows down the workflow and creates bottlenecks.
Lack of accountability: When teammates do not take respons...
I would respectfully discuss my concerns with my manager and propose alternative solutions.
Initiate a conversation with the manager to understand their perspective
Express concerns and provide logical reasoning for alternative approaches
Present data or examples to support the proposed solutions
Maintain a respectful and professional attitude throughout the discussion
Seeking new opportunities to further develop my skills and contribute to a more challenging and dynamic environment.
Looking for new challenges and growth opportunities
Seeking a more dynamic and innovative work environment
Wanting to work on a wider range of projects and gain diverse experience
Desire to contribute to a company with a stronger focus on data science
Exploring opportunities for career advancement and profess
I am looking for new opportunities or change because I want to further develop my skills and contribute to a dynamic and innovative team.
Seeking new challenges and growth opportunities
Interested in working with a diverse and talented team
Want to contribute to innovative projects and make a meaningful impact
Desire to expand knowledge and skills in data science
Looking for a company culture that aligns with my values and
Exploring more offers to make an informed decision and maximize career growth.
To evaluate the best fit for my skills, interests, and long-term goals.
To negotiate better compensation and benefits.
To gain exposure to different industries, technologies, and projects.
To expand professional network and opportunities.
To ensure job security and minimize risks.
To have a backup option in case one offer falls through.
To explore ...
I applied via Recruitment Consulltant and was interviewed before Feb 2023. There were 2 interview rounds.
2 question
1. Find the highest stock proce profit in a day.
2. To rollup and aggregate claims at covergae level.
Top trending discussions
I applied via Naukri.com and was interviewed in Aug 2024. There was 1 interview round.
Supervised learning is a type of machine learning where the model is trained on labeled data to make predictions or decisions.
Uses labeled data for training
Predicts outcomes based on input features
Examples include regression and classification algorithms
Unsupervised learning is a type of machine learning where the model is trained on unlabeled data without any predefined output labels.
No predefined output labels are provided for the training data
The model must find patterns and relationships in the data on its own
Common techniques include clustering and dimensionality reduction
Examples: K-means clustering, Principal Component Analysis (PCA)
posted on 17 Mar 2024
I applied via Naukri.com and was interviewed in Aug 2024. There was 1 interview round.
Supervised learning is a type of machine learning where the model is trained on labeled data to make predictions or decisions.
Uses labeled data for training
Predicts outcomes based on input features
Examples include regression and classification algorithms
Unsupervised learning is a type of machine learning where the model is trained on unlabeled data without any predefined output labels.
No predefined output labels are provided for the training data
The model must find patterns and relationships in the data on its own
Common techniques include clustering and dimensionality reduction
Examples: K-means clustering, Principal Component Analysis (PCA)
LSTM RNN is a type of RNN that can learn long-term dependencies, while simple RNN struggles with vanishing/exploding gradients.
LSTM RNN has more complex architecture with memory cells, input, forget, and output gates.
Simple RNN has a single tanh activation function and suffers from vanishing/exploding gradients.
LSTM RNN is better at capturing long-term dependencies in sequences.
Simple RNN is simpler but struggles with
Lasso regression is a type of linear regression that uses L1 regularization to prevent overfitting by adding a penalty term to the loss function.
Lasso regression helps in feature selection by shrinking the coefficients of less important features to zero.
It is particularly useful when dealing with high-dimensional data where the number of features is much larger than the number of samples.
The regularization parameter in...
Fine tuning a LLM model involves adjusting hyperparameters to improve performance.
Perform grid search or random search to find optimal hyperparameters
Use cross-validation to evaluate different hyperparameter combinations
Regularize the model to prevent overfitting
Adjust learning rate and batch size for better convergence
Consider using techniques like early stopping to prevent overfitting
posted on 13 Aug 2024
I applied via Recruitment Consulltant and was interviewed in Feb 2024. There was 1 interview round.
Deep learning is used over statistical models for complex, non-linear relationships in data.
Deep learning can automatically learn hierarchical representations of data, capturing intricate patterns and relationships.
Statistical models may struggle with high-dimensional data or non-linear relationships, where deep learning excels.
Deep learning can handle unstructured data like images, audio, and text more effectively tha...
XGB is better than RF due to its ability to handle complex relationships and optimize performance.
XGB uses gradient boosting which allows for better handling of complex relationships compared to RF
XGB optimizes performance by using regularization techniques to prevent overfitting
XGB is faster and more efficient in training compared to RF
XGB allows for parallel processing which can speed up computation
XGB has been shown...
posted on 17 Mar 2024
based on 3 reviews
Rating in categories
Senior Software Engineer
230
salaries
| ₹9.2 L/yr - ₹32 L/yr |
Technical Lead
198
salaries
| ₹16 L/yr - ₹46 L/yr |
Software Engineer
113
salaries
| ₹6 L/yr - ₹17 L/yr |
Data Scientist
42
salaries
| ₹8.5 L/yr - ₹25.3 L/yr |
Senior Analyst
42
salaries
| ₹6.5 L/yr - ₹23 L/yr |
AIG
Allianz
AXA
Zurich Insurance