Sensitivity in confusion matrix is
Anonymous Quiz
18%
True negative rate
39%
True positive rate
34%
Both A and B
9%
None of these
In regression analysis, if the independent variable is measured in kilograms, the dependent variable
Anonymous Quiz
33%
must also be in kilograms
20%
must be in some unit of weight
7%
cannot be in kilograms
40%
can be any units
Overfitting is a major problem for neural networks. Which of the following can help prevent overfitting?
Anonymous Quiz
8%
Retraining on the same data many times
23%
Using a larger learning rate for Backpropagation
52%
Dropping random neurons in each iteration of Backpropagation
17%
Training until you get the smallest training error
Regression line can be drawn in which of the following plots
Anonymous Quiz
7%
Pair plot
36%
Regression plot
4%
Joint plot
52%
All of the above
Which of the following is statements is false
Anonymous Quiz
24%
Boosting combines weak classifiers to output a strong classifier
24%
Adaboost is a generative model as it can generate classifier
32%
Boosting can only improve performance if weak learners can predict better than random chance
21%
None of these
High entropy means that the partitions in classification
Anonymous Quiz
27%
Pure
44%
Not pure
18%
Useful
12%
Useless
Which of the following is a way to avoid local minima?
Anonymous Quiz
19%
Increase the learning rate
23%
Use momentum and Adaptive learning
12%
Add some noise while updating weights
46%
All the above
Curated papers, articles, and blogs on data science & machine learning in production.
https://github.com/eugeneyan/applied-ml
https://github.com/eugeneyan/applied-ml
GitHub
GitHub - eugeneyan/applied-ml: 📚 Papers & tech blogs by companies sharing their work on data science & machine learning in production.
📚 Papers & tech blogs by companies sharing their work on data science & machine learning in production. - eugeneyan/applied-ml
Which of the following is FALSE about Batch Normalization?
Anonymous Quiz
31%
It may lead to overfitting
20%
It solves dying ReLU problem
22%
It introduces regularization in the network
27%
It solves vanishing and exploding gradient issues
Tableau Cheatsheet_ Follow Dr. AngShu (@drangshu) for more.pdf
165.2 KB
Emailing Tableau Cheatsheet
More than 70+ cheatsheets related to data science, machine learning, deep learning and computer vision in python and R programming language.
https://www.theinsaneapp.com/2020/12/machine-learning-and-data-science-cheat-sheets-pdf.html
https://www.theinsaneapp.com/2020/12/machine-learning-and-data-science-cheat-sheets-pdf.html
In following type of feature selection method we start with empty feature set
Anonymous Quiz
44%
Forward feature selection
19%
Backward feature selection
29%
Both A and B
9%
None of these
Handwritten notes of machine learning
https://www.linkedin.com/posts/data-spoof_handwritten-notes-on-machine-learning-activity-6773890970741202944-rG2R
https://www.linkedin.com/posts/data-spoof_handwritten-notes-on-machine-learning-activity-6773890970741202944-rG2R
Linkedin
DataSpoof on LinkedIn: Handwritten notes on Machine learning
Handwritten notes on Machine learning.
Like ❤️, comment, and share the post with your friends.
Follow DataSpoof for daily educational contents and study…
Like ❤️, comment, and share the post with your friends.
Follow DataSpoof for daily educational contents and study…
Which of the following is well suited for perceptual tasks?
Anonymous Quiz
24%
Recurrent neural network
32%
Reinforcement learning
30%
Convolutional neural network
14%
Feed forward neural network
Which of the following is a good test dataset characteristic?
Anonymous Quiz
12%
large enough to yield meaningful results
23%
is representative of the dataset as a whole
62%
Both A and B
2%
None of the above
Which of the following is true about Residuals?
Anonymous Quiz
62%
Lower is better
12%
Higher is better
17%
Both A and B
8%
None of these
Classification problems are distinguished from estimation problems in that
Anonymous Quiz
13%
classification problems require the output attribute to be numeric.
64%
classification problems require the output attribute to be categorical
8%
classification problems do not allow an output attribute.
14%
classification problems are designed to predict future outcome.